Loading ...
Sorry, an error occurred while loading the content.

Listening and engineers

Expand Messages
  • Robert
    I think we are somehow not getting the message of Staffelt s article, probably because I have not made it clear. According to his article, NO system that
    Message 1 of 24 , Oct 5, 2010
    • 0 Attachment
      I think we are somehow not getting the message of
      Staffelt's article, probably because I have not made
      it clear.

      According to his article, NO system that measures
      response in the abstract(in the absence of the head)
      is going to work right.

      His claim is that the thing that needs EQ(though NOT to flat)
      is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
      This can only be obtained by using dummy head measurement
      or even better measurement on your own head.
      Alternatively--and this it the REAL POINT--you should do this
      by listening.

      Almost everyone is ultimately disappointed in almost all the
      automatic systems, except the ones which have detailed user
      control.

      YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
      Automatic roughs it out, but after that, you need to fine tune by ear.

      Engineers do not like this, philoisophically. It seems unscientific
      and "subjective" to them. But to me it makes perfect sense and
      also fits with my experience.

      I have really discovered that I do not like any automatic system
      for the finishing touches. ALL the usual models of how we hear sound in rooms are wrong, except the ones that are so complicated that one cannot use them or the ones that measure with your head in place.
      They have to be wrong, actually because of how complicated things are. A correct model would be a lot more complicated than a single frequency varying time window for example.

      But none of those latter kind(dummy head or your own head) are offered for sale.
      Every system on the market measures the sound in the absence of the head. And almost all of them try to deal with things by using time windows. According to Staffelt this won't work and in my experience it does not work, not exactly.

      None of this automatic stuff really works very well above the bass. But this is not to say that SOMETHING does not work well above the bass. As I used to say to GF, since you can change the sound in the higher frequencies you are almost bound to be able to make it better!

      The problem is the same as audio has always had. Engineers like engineering solutions, and solutions that involve human experience as an evaluation tool make engineers nervous, in general.

      I think it makes a whole lot of sense simply to get pink noise to sound right. Of course you have somehow to find out what right means, as TM suggested. But you can experiment and learn to hear what is right quite easily. Just try a lot of EQs both on pink noise and on recordings that you think ought to be good and set up something that makes them sound right in some intuitive sense.

      This could take some time! But it beats messing around trying to do the same thing by buying different equipment items or modifying your room. And it is a lot cheaper too.

      Of course one needs some sort of half decent listening room to begin with. But that is not so hard to arrange.

      I think the big failing here is human--people have not worked with user controlled EQ carefully enough. Talk about laziness! People would apparently rather buy stuff than work even a little.
      This is craziness to my mind. Anyway, once you get into it, it is FUN
      to play with EQ.

      I find it really intriguing to try to adjust one speaker to sound like another. Interesting to try!

      And don't forget that NO speaker, essentially , is as smooth as it ought to be. Just look at those NRC measurements. Forget questions of what radiation pattern ought to be for the moment and just look at one axis response. ALmost NO speaker is as smooth as it ought to be.
      It very seldom happens. And when adds in the effects of off axis--well, I think it is fair to say that it NEVER happens, that it is as smooth as it ought to be(independently of what "flat " means, everyone agrees that things ought to be really smooth--but they are not, not in detail. Even getting totally smooth 1/3 octave is quite unusual without EQ).

      Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).

      Engineers don't like this idea and most of them do not trust the customers either. But it is your music--and your head! whose transfer function you listen to.

      REG
    • uohh
      I m confused! Since we have very different head shapes and ear canal anatomies, are you suggesting that we need to EQ on an individual basis? And since
      Message 2 of 24 , Oct 5, 2010
      • 0 Attachment
        I'm confused! Since we have very different head shapes and ear canal anatomies, are you suggesting that we need to EQ on an individual basis?

        And since recordings were not made with microphones inside dummy heads, it follows that the recording records the sound as in the free space the mic "hears" without the influence of a human head or ear canal. So what is then the reasoning supporting the need to measure my audio system using mics inside dummy heads, if we are not trying to reproduce what is being heard by my ear in the recording venue?

        Did I mis-understand what you were saying?

        Joseph


        --- In regsaudioforum@yahoogroups.com, "Robert" <regonaudio@...> wrote:
        >
        >
        >
        > I think we are somehow not getting the message of
        > Staffelt's article, probably because I have not made
        > it clear.
        >
        > According to his article, NO system that measures
        > response in the abstract(in the absence of the head)
        > is going to work right.
        >
        > His claim is that the thing that needs EQ(though NOT to flat)
        > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
        > This can only be obtained by using dummy head measurement
        > or even better measurement on your own head.
        > Alternatively--and this it the REAL POINT--you should do this
        > by listening.
        >
        > Almost everyone is ultimately disappointed in almost all the
        > automatic systems, except the ones which have detailed user
        > control.
        >
        > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
        > Automatic roughs it out, but after that, you need to fine tune by ear.
        >
        > Engineers do not like this, philoisophically. It seems unscientific
        > and "subjective" to them. But to me it makes perfect sense and
        > also fits with my experience.
        >
        > I have really discovered that I do not like any automatic system
        > for the finishing touches. ALL the usual models of how we hear sound in rooms are wrong, except the ones that are so complicated that one cannot use them or the ones that measure with your head in place.
        > They have to be wrong, actually because of how complicated things are. A correct model would be a lot more complicated than a single frequency varying time window for example.
        >
        > But none of those latter kind(dummy head or your own head) are offered for sale.
        > Every system on the market measures the sound in the absence of the head. And almost all of them try to deal with things by using time windows. According to Staffelt this won't work and in my experience it does not work, not exactly.
        >
        > None of this automatic stuff really works very well above the bass. But this is not to say that SOMETHING does not work well above the bass. As I used to say to GF, since you can change the sound in the higher frequencies you are almost bound to be able to make it better!
        >
        > The problem is the same as audio has always had. Engineers like engineering solutions, and solutions that involve human experience as an evaluation tool make engineers nervous, in general.
        >
        > I think it makes a whole lot of sense simply to get pink noise to sound right. Of course you have somehow to find out what right means, as TM suggested. But you can experiment and learn to hear what is right quite easily. Just try a lot of EQs both on pink noise and on recordings that you think ought to be good and set up something that makes them sound right in some intuitive sense.
        >
        > This could take some time! But it beats messing around trying to do the same thing by buying different equipment items or modifying your room. And it is a lot cheaper too.
        >
        > Of course one needs some sort of half decent listening room to begin with. But that is not so hard to arrange.
        >
        > I think the big failing here is human--people have not worked with user controlled EQ carefully enough. Talk about laziness! People would apparently rather buy stuff than work even a little.
        > This is craziness to my mind. Anyway, once you get into it, it is FUN
        > to play with EQ.
        >
        > I find it really intriguing to try to adjust one speaker to sound like another. Interesting to try!
        >
        > And don't forget that NO speaker, essentially , is as smooth as it ought to be. Just look at those NRC measurements. Forget questions of what radiation pattern ought to be for the moment and just look at one axis response. ALmost NO speaker is as smooth as it ought to be.
        > It very seldom happens. And when adds in the effects of off axis--well, I think it is fair to say that it NEVER happens, that it is as smooth as it ought to be(independently of what "flat " means, everyone agrees that things ought to be really smooth--but they are not, not in detail. Even getting totally smooth 1/3 octave is quite unusual without EQ).
        >
        > Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).
        >
        > Engineers don't like this idea and most of them do not trust the customers either. But it is your music--and your head! whose transfer function you listen to.
        >
        > REG
        >
      • Ken Holder
        ... Well, that made me laugh out loud. Are there measurements microphones that fit in the ear canal? Can one of those Styrofoam wig-heads be used as a dummy
        Message 3 of 24 , Oct 5, 2010
        • 0 Attachment
          At 06:25 PM 10/5/2010, Robert wrote:

          >But none of those latter kind(dummy head or your own head) are
          >offered for sale.


          Well, that made me laugh out loud.

          Are there measurements microphones that fit in the ear
          canal?

          Can one of those Styrofoam wig-heads be used as a dummy
          head?

          Is a third-octave unit with knobs easier to fiddle with
          than one with a LCD readout with up-down level buttons
          and left-right band selection buttons (the Behringer
          drives me nutty (-er than I already am I mean).)?

          Still, Acourate (tm) has improved the sound of my system
          to an astonishing degree. I'm still just blown away by
          how beautiful the best recordings sound. And even the
          second-best aren't bad. Heck, I listened to Stokowski's
          1939 recording of the Shostakovich 5th symphony last
          evening and IT even sounded pretty darn good. (And the
          Academy of Music sounded the same in 1939 as it did in
          1999 -- those recordings with more reverb must have
          gotten it from the famous CBS stairwell enhancement
          system!)

          Ken Holder
          Just a Poor, Old, Simple, Country-Living, Music-Lover
        • Uli Brueggemann
          Robert, I really have problems to understand what you mean. BTW where can I get this Staffelt article? I have not found it. Imagine a musician playing in your
          Message 4 of 24 , Oct 6, 2010
          • 0 Attachment
            Robert,

            I really have problems to understand what you mean.
            BTW where can I get this Staffelt article? I have not found it.

            Imagine a musician playing in your room. The sound that arrives to several listeners' ear canals is independent of the ear canal. The sound will be treated by the ear canal and the auditory system and the brain of the listener.But the arrival of the sound is independent of the head.

            Imagine you record the musician and you play his music by a loudspeaker. If the sound arriving at the ear canal would be identical to the the sound before then a listener would not have chance to distinguish just by listening.

            To achieve such an ideal situation of course the recording must be perfect. Imagine we have such a perfect recording then of course we also have to conclude a perfectness of the playback. But only in the direction that the sound arriving at the ear canal is identical.

            Ok, back to the real world. We have recordings and they are more or less perfect. In most cases less perfect. But we cannot change them, we can just play them.
            Assuming that the recording is what it is we may search for a perfect playback. A signal fed into the speaker should arrive at the ear canal unchanged.

            But of course the sound is influenced by the speaker and by the room. And now we can start to discuss how to improve the sound.

            What happens if the sound arriving at the nose of the listener is ok? What happens if the sound arriving at the top of the listeners head is ok?
            I post the hypothesis that if a sound is ok in a certain radius around the listeners head it is also ok at the ear canal. Thus it is not necessary to measure with a mic inside the ear. In this case the sound at the microphone is already bent by the ears. And correcting the sound according to the in-ear-recording will lead to a double treatment of the sound as the corrected sound will be bent again by the listening ear.

            This all does not make sense to me.

            I recommend you to think about wavelet transforms and the Stockwell Transform. In general you will detect that you get an unsharp picture of the behaviour in frequency and time domain. We are back to the Uncertainty Principle of Heisenberg (established for a different purpose but still applicable). If we have no maths to describe both frequency/phase and time behaviour then we must conclude that our ear/brain is also not capable to analyse this perfectly.

            When we listen we recognize both frequency and time aspects in the music. But we do not get a sharp picture. A Stockwell Transform shows a smooth frequency response. Much different compared to a steady state Fourier Transform of a room pulse response. Interestingly we can also get a smoothed frequency response when we apply a frequency dependent windowing and it will be very close to the frequency response of a Stockwell Transform. This means that by a frequency dependent windowing (btw also with a 1/3 octave analysis or other several smoothing methods) we get a picture close to what we perceive. Of course we may discuss about the best window sizes or the size of n in the 1/n-octave analysis or the proper Erb scale ...

            I fully agree that a one button automatic correction is comfortable for the user but typically will not result in an optimal correction. Thus I prefer to see the measurements, to see what happens with a correction and to be able to finetune a correction. This gives a good understanding what's going on.

            A dummy head is the wrong way.

            Best regards

            Uli




            On Wed, Oct 6, 2010 at 3:25 AM, Robert <regonaudio@...> wrote:
             



            I think we are somehow not getting the message of
            Staffelt's article, probably because I have not made
            it clear.

            According to his article, NO system that measures
            response in the abstract(in the absence of the head)
            is going to work right.

            His claim is that the thing that needs EQ(though NOT to flat)
            is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
            This can only be obtained by using dummy head measurement
            or even better measurement on your own head.
            Alternatively--and this it the REAL POINT--you should do this
            by listening.

            Almost everyone is ultimately disappointed in almost all the
            automatic systems, except the ones which have detailed user
            control.

            YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
            Automatic roughs it out, but after that, you need to fine tune by ear.

            Engineers do not like this, philoisophically. It seems unscientific
            and "subjective" to them. But to me it makes perfect sense and
            also fits with my experience.

            I have really discovered that I do not like any automatic system
            for the finishing touches. ALL the usual models of how we hear sound in rooms are wrong, except the ones that are so complicated that one cannot use them or the ones that measure with your head in place.
            They have to be wrong, actually because of how complicated things are. A correct model would be a lot more complicated than a single frequency varying time window for example.

            But none of those latter kind(dummy head or your own head) are offered for sale.
            Every system on the market measures the sound in the absence of the head. And almost all of them try to deal with things by using time windows. According to Staffelt this won't work and in my experience it does not work, not exactly.

            None of this automatic stuff really works very well above the bass. But this is not to say that SOMETHING does not work well above the bass. As I used to say to GF, since you can change the sound in the higher frequencies you are almost bound to be able to make it better!

            The problem is the same as audio has always had. Engineers like engineering solutions, and solutions that involve human experience as an evaluation tool make engineers nervous, in general.

            I think it makes a whole lot of sense simply to get pink noise to sound right. Of course you have somehow to find out what right means, as TM suggested. But you can experiment and learn to hear what is right quite easily. Just try a lot of EQs both on pink noise and on recordings that you think ought to be good and set up something that makes them sound right in some intuitive sense.

            This could take some time! But it beats messing around trying to do the same thing by buying different equipment items or modifying your room. And it is a lot cheaper too.

            Of course one needs some sort of half decent listening room to begin with. But that is not so hard to arrange.

            I think the big failing here is human--people have not worked with user controlled EQ carefully enough. Talk about laziness! People would apparently rather buy stuff than work even a little.
            This is craziness to my mind. Anyway, once you get into it, it is FUN
            to play with EQ.

            I find it really intriguing to try to adjust one speaker to sound like another. Interesting to try!

            And don't forget that NO speaker, essentially , is as smooth as it ought to be. Just look at those NRC measurements. Forget questions of what radiation pattern ought to be for the moment and just look at one axis response. ALmost NO speaker is as smooth as it ought to be.
            It very seldom happens. And when adds in the effects of off axis--well, I think it is fair to say that it NEVER happens, that it is as smooth as it ought to be(independently of what "flat " means, everyone agrees that things ought to be really smooth--but they are not, not in detail. Even getting totally smooth 1/3 octave is quite unusual without EQ).

            Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).

            Engineers don't like this idea and most of them do not trust the customers either. But it is your music--and your head! whose transfer function you listen to.

            REG


          • HM
            One of the companies using Dummy Head microphones for measurement is http://www.head-acoustics.de/eng/index.htm Not to forget Brüel&Kjaer, both companies
            Message 5 of 24 , Oct 6, 2010
            • 0 Attachment
              One of the companies using Dummy Head microphones for measurement is http://www.head-acoustics.de/eng/index.htm
              Not to forget Brüel&Kjaer, both companies are well renowned in measurment.
              Neumann, now a branch of Sennheiser (Oskar dummy Head and binaural mics for individuals like MKE2002) had their KU80, KU81 and KU100, the first with the mic inside the artificial ear canal, the latter with a mic close to the outer ear. This made recordings more versatile.
              Here are pictures and a list:
              http://www.green-sonic.net/deutsch/kunstkopf.html

              EMI central research lab developed their Sensaura recording technique based on artificial head mics and DSP-correction of the HRTF to make it playable via speakers. This created some coloration, but astounding soundstage with the RAF band marching your room nearly 90° on either side.

              I have done measurement and correction with TacT RCS, based on 2 mic positions, one for each ear. The effect is a very transparent soundstage.
              There are lots of combfilter effects if one moves the head only slightly, which makes the system quite useless.

              Like Uli, I dont believe Staffelt.
              Blauert (Spatial Hearing) and Theile have written a lot about this subject, some articles are available from the Neuman website.

              If the mic is placed at the ear, a headphone placed at the same position is best.
              The signal travels the ear canal then once which is part of natural listening.
              If the mic is placed inside the canal for a recording, the sound would pass the canal for a second time when listening.

              When listening with in-ear headphones, Siegfried Linkwitz has corrected the earcanal resonance because the canal is closed at either end this is relevant.
              With the open ear the canal ends in a kind of horn shape and this IMO means coupling to the air without reflexion of the canals resonance frequency.
              BR HM



              --- In regsaudioforum@yahoogroups.com, "Robert" <regonaudio@...> wrote:
              >
              >
              >
              > I think we are somehow not getting the message of
              > Staffelt's article, probably because I have not made
              > it clear.
              >
              > According to his article, NO system that measures
              > response in the abstract(in the absence of the head)
              > is going to work right.
              >
              > His claim is that the thing that needs EQ(though NOT to flat)
              > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
              > This can only be obtained by using dummy head measurement
              > or even better measurement on your own head.
              > Alternatively--and this it the REAL POINT--you should do this
              > by listening.
              >
              > Almost everyone is ultimately disappointed in almost all the
              > automatic systems, except the ones which have detailed user
              > control.
              >
              > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
              > Automatic roughs it out, but after that, you need to fine tune by ear.
              >
              > Engineers do not like this, philoisophically. It seems unscientific
              > and "subjective" to them. But to me it makes perfect sense and
              > also fits with my experience.
              >
              > I have really discovered that I do not like any automatic system
              > for the finishing touches. ALL the usual models of how we hear sound in rooms are wrong, except the ones that are so complicated that one cannot use them or the ones that measure with your head in place.
              > They have to be wrong, actually because of how complicated things are. A correct model would be a lot more complicated than a single frequency varying time window for example.
              >
              > But none of those latter kind(dummy head or your own head) are offered for sale.
              > Every system on the market measures the sound in the absence of the head. And almost all of them try to deal with things by using time windows. According to Staffelt this won't work and in my experience it does not work, not exactly.
              >
              > None of this automatic stuff really works very well above the bass. But this is not to say that SOMETHING does not work well above the bass. As I used to say to GF, since you can change the sound in the higher frequencies you are almost bound to be able to make it better!
              >
              > The problem is the same as audio has always had. Engineers like engineering solutions, and solutions that involve human experience as an evaluation tool make engineers nervous, in general.
              >
              > I think it makes a whole lot of sense simply to get pink noise to sound right. Of course you have somehow to find out what right means, as TM suggested. But you can experiment and learn to hear what is right quite easily. Just try a lot of EQs both on pink noise and on recordings that you think ought to be good and set up something that makes them sound right in some intuitive sense.
              >
              > This could take some time! But it beats messing around trying to do the same thing by buying different equipment items or modifying your room. And it is a lot cheaper too.
              >
              > Of course one needs some sort of half decent listening room to begin with. But that is not so hard to arrange.
              >
              > I think the big failing here is human--people have not worked with user controlled EQ carefully enough. Talk about laziness! People would apparently rather buy stuff than work even a little.
              > This is craziness to my mind. Anyway, once you get into it, it is FUN
              > to play with EQ.
              >
              > I find it really intriguing to try to adjust one speaker to sound like another. Interesting to try!
              >
              > And don't forget that NO speaker, essentially , is as smooth as it ought to be. Just look at those NRC measurements. Forget questions of what radiation pattern ought to be for the moment and just look at one axis response. ALmost NO speaker is as smooth as it ought to be.
              > It very seldom happens. And when adds in the effects of off axis--well, I think it is fair to say that it NEVER happens, that it is as smooth as it ought to be(independently of what "flat " means, everyone agrees that things ought to be really smooth--but they are not, not in detail. Even getting totally smooth 1/3 octave is quite unusual without EQ).
              >
              > Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).
              >
              > Engineers don't like this idea and most of them do not trust the customers either. But it is your music--and your head! whose transfer function you listen to.
              >
              > REG
              >
            • plinque_fish
              It s worth looking at how Autocal works. It is automatic, but after initial calibration the user can make any further eq adjustments they want to via the
              Message 6 of 24 , Oct 6, 2010
              • 0 Attachment
                It's worth looking at how Autocal works.
                It is automatic, but after initial calibration the user can make any further eq adjustments they want to via the computer, and then store those adjustments into each speaker so that in use the speakers do not have to be connected to an external PC. The software is calibrated (as far as I can tell) for the response of individual speakers and mics. I imagine it is would be very easy to eq the bass alone if that's all you wanted to do.


                > Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).
              • Tom Mallin
                Uli, I think it is quite obvious that the human head plays a major role in how each individual in a listening room hears sound played back from speakers. The
                Message 7 of 24 , Oct 6, 2010
                • 0 Attachment
                  Uli, I think it is quite obvious that the human head plays a major role in how each individual in a listening room hears sound played back from speakers. The head shadowing effect and time delays between ears are very real for each listener. See http://www.regonaudio.com/Directional%20Hearing%20How%20To%20Listen%20to%20Stereo.htm

                  The reason that headphone listening to recordings made to be played back through speakers is ultimately unrealistic is precisely because headphones eliminate the effect of the human head on how we hear sound. They eliminate head shadowing, time delay effects between ears, and also result in sound being injected into the ear canal directly from the side rather than from some angle toward the front of the head, altering the perceived frequency response.

                  But, as REG said, the real point of what he is saying is that, based on REG's reading of the Staffelt research, the human ear/brain's subjective judgment should be relied upon above the bass in adjusting EQ.

                  >>> Uli Brueggemann <uli.brueggemann@...> 10/6/2010 2:16 AM >>>


                  Robert,

                  I really have problems to understand what you mean.
                  BTW where can I get this Staffelt article? I have not found it.

                  Imagine a musician playing in your room. The sound that arrives to several listeners' ear canals is independent of the ear canal. The sound will be treated by the ear canal and the auditory system and the brain of the listener.But the arrival of the sound is independent of the head.

                  Imagine you record the musician and you play his music by a loudspeaker. If the sound arriving at the ear canal would be identical to the the sound before then a listener would not have chance to distinguish just by listening.

                  To achieve such an ideal situation of course the recording must be perfect. Imagine we have such a perfect recording then of course we also have to conclude a perfectness of the playback. But only in the direction that the sound arriving at the ear canal is identical.

                  Ok, back to the real world. We have recordings and they are more or less perfect. In most cases less perfect. But we cannot change them, we can just play them.
                  Assuming that the recording is what it is we may search for a perfect playback. A signal fed into the speaker should arrive at the ear canal unchanged.

                  But of course the sound is influenced by the speaker and by the room. And now we can start to discuss how to improve the sound.

                  What happens if the sound arriving at the nose of the listener is ok? What happens if the sound arriving at the top of the listeners head is ok?
                  I post the hypothesis that if a sound is ok in a certain radius around the listeners head it is also ok at the ear canal. Thus it is not necessary to measure with a mic inside the ear. In this case the sound at the microphone is already bent by the ears. And correcting the sound according to the in-ear-recording will lead to a double treatment of the sound as the corrected sound will be bent again by the listening ear.

                  This all does not make sense to me.

                  I recommend you to think about wavelet transforms and the Stockwell Transform. In general you will detect that you get an unsharp picture of the behaviour in frequency and time domain. We are back to the Uncertainty Principle of Heisenberg (established for a different purpose but still applicable). If we have no maths to describe both frequency/phase and time behaviour then we must conclude that our ear/brain is also not capable to analyse this perfectly.

                  When we listen we recognize both frequency and time aspects in the music. But we do not get a sharp picture. A Stockwell Transform shows a smooth frequency response. Much different compared to a steady state Fourier Transform of a room pulse response. Interestingly we can also get a smoothed frequency response when we apply a frequency dependent windowing and it will be very close to the frequency response of a Stockwell Transform. This means that by a frequency dependent windowing (btw also with a 1/3 octave analysis or other several smoothing methods) we get a picture close to what we perceive. Of course we may discuss about the best window sizes or the size of n in the 1/n-octave analysis or the proper Erb scale ...

                  I fully agree that a one button automatic correction is comfortable for the user but typically will not result in an optimal correction. Thus I prefer to see the measurements, to see what happens with a correction and to be able to finetune a correction. This gives a good understanding what's going on.

                  A dummy head is the wrong way.

                  Best regards

                  Uli
                • mangmeng
                  http://www.etymotic.com/ephp/er4-wrfy.aspx Is this info of any use in the discussion. Yip
                  Message 8 of 24 , Oct 6, 2010
                  • 0 Attachment
                    http://www.etymotic.com/ephp/er4-wrfy.aspx

                    Is this info of any use in the discussion.

                    Yip

                    --- In regsaudioforum@yahoogroups.com, "HM" <hmartinburm@...> wrote:
                    >
                    > One of the companies using Dummy Head microphones for measurement is http://www.head-acoustics.de/eng/index.htm
                    > Not to forget Brüel&Kjaer, both companies are well renowned in measurment.
                    > Neumann, now a branch of Sennheiser (Oskar dummy Head and binaural mics for individuals like MKE2002) had their KU80, KU81 and KU100, the first with the mic inside the artificial ear canal, the latter with a mic close to the outer ear. This made recordings more versatile.
                    > Here are pictures and a list:
                    > http://www.green-sonic.net/deutsch/kunstkopf.html
                    >
                    > EMI central research lab developed their Sensaura recording technique based on artificial head mics and DSP-correction of the HRTF to make it playable via speakers. This created some coloration, but astounding soundstage with the RAF band marching your room nearly 90° on either side.
                    >
                    > I have done measurement and correction with TacT RCS, based on 2 mic positions, one for each ear. The effect is a very transparent soundstage.
                    > There are lots of combfilter effects if one moves the head only slightly, which makes the system quite useless.
                    >
                    > Like Uli, I dont believe Staffelt.
                    > Blauert (Spatial Hearing) and Theile have written a lot about this subject, some articles are available from the Neuman website.
                    >
                    > If the mic is placed at the ear, a headphone placed at the same position is best.
                    > The signal travels the ear canal then once which is part of natural listening.
                    > If the mic is placed inside the canal for a recording, the sound would pass the canal for a second time when listening.
                    >
                    > When listening with in-ear headphones, Siegfried Linkwitz has corrected the earcanal resonance because the canal is closed at either end this is relevant.
                    > With the open ear the canal ends in a kind of horn shape and this IMO means coupling to the air without reflexion of the canals resonance frequency.
                    > BR HM
                    >
                    >
                    >
                    > --- In regsaudioforum@yahoogroups.com, "Robert" <regonaudio@> wrote:
                    > >
                    > >
                    > >
                    > > I think we are somehow not getting the message of
                    > > Staffelt's article, probably because I have not made
                    > > it clear.
                    > >
                    > > According to his article, NO system that measures
                    > > response in the abstract(in the absence of the head)
                    > > is going to work right.
                    > >
                    > > His claim is that the thing that needs EQ(though NOT to flat)
                    > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                    > > This can only be obtained by using dummy head measurement
                    > > or even better measurement on your own head.
                    > > Alternatively--and this it the REAL POINT--you should do this
                    > > by listening.
                    > >
                    > > Almost everyone is ultimately disappointed in almost all the
                    > > automatic systems, except the ones which have detailed user
                    > > control.
                    > >
                    > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                    > > Automatic roughs it out, but after that, you need to fine tune by ear.
                    > >
                    > > Engineers do not like this, philoisophically. It seems unscientific
                    > > and "subjective" to them. But to me it makes perfect sense and
                    > > also fits with my experience.
                    > >
                    > > I have really discovered that I do not like any automatic system
                    > > for the finishing touches. ALL the usual models of how we hear sound in rooms are wrong, except the ones that are so complicated that one cannot use them or the ones that measure with your head in place.
                    > > They have to be wrong, actually because of how complicated things are. A correct model would be a lot more complicated than a single frequency varying time window for example.
                    > >
                    > > But none of those latter kind(dummy head or your own head) are offered for sale.
                    > > Every system on the market measures the sound in the absence of the head. And almost all of them try to deal with things by using time windows. According to Staffelt this won't work and in my experience it does not work, not exactly.
                    > >
                    > > None of this automatic stuff really works very well above the bass. But this is not to say that SOMETHING does not work well above the bass. As I used to say to GF, since you can change the sound in the higher frequencies you are almost bound to be able to make it better!
                    > >
                    > > The problem is the same as audio has always had. Engineers like engineering solutions, and solutions that involve human experience as an evaluation tool make engineers nervous, in general.
                    > >
                    > > I think it makes a whole lot of sense simply to get pink noise to sound right. Of course you have somehow to find out what right means, as TM suggested. But you can experiment and learn to hear what is right quite easily. Just try a lot of EQs both on pink noise and on recordings that you think ought to be good and set up something that makes them sound right in some intuitive sense.
                    > >
                    > > This could take some time! But it beats messing around trying to do the same thing by buying different equipment items or modifying your room. And it is a lot cheaper too.
                    > >
                    > > Of course one needs some sort of half decent listening room to begin with. But that is not so hard to arrange.
                    > >
                    > > I think the big failing here is human--people have not worked with user controlled EQ carefully enough. Talk about laziness! People would apparently rather buy stuff than work even a little.
                    > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                    > > to play with EQ.
                    > >
                    > > I find it really intriguing to try to adjust one speaker to sound like another. Interesting to try!
                    > >
                    > > And don't forget that NO speaker, essentially , is as smooth as it ought to be. Just look at those NRC measurements. Forget questions of what radiation pattern ought to be for the moment and just look at one axis response. ALmost NO speaker is as smooth as it ought to be.
                    > > It very seldom happens. And when adds in the effects of off axis--well, I think it is fair to say that it NEVER happens, that it is as smooth as it ought to be(independently of what "flat " means, everyone agrees that things ought to be really smooth--but they are not, not in detail. Even getting totally smooth 1/3 octave is quite unusual without EQ).
                    > >
                    > > Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).
                    > >
                    > > Engineers don't like this idea and most of them do not trust the customers either. But it is your music--and your head! whose transfer function you listen to.
                    > >
                    > > REG
                    > >
                    >
                  • mangmeng
                    http://www.etymotic.com/ephp/er4-hwmra.aspx some more info from the ER site. Yip
                    Message 9 of 24 , Oct 6, 2010
                    • 0 Attachment
                      http://www.etymotic.com/ephp/er4-hwmra.aspx

                      some more info from the ER site.

                      Yip

                      --- In regsaudioforum@yahoogroups.com, "mangmeng" <yipmangmeng@...> wrote:
                      >
                      > http://www.etymotic.com/ephp/er4-wrfy.aspx
                      >
                      > Is this info of any use in the discussion.
                      >
                      > Yip
                      >
                      > --- In regsaudioforum@yahoogroups.com, "HM" <hmartinburm@> wrote:
                      > >
                      > > One of the companies using Dummy Head microphones for measurement is http://www.head-acoustics.de/eng/index.htm
                      > > Not to forget Brüel&Kjaer, both companies are well renowned in measurment.
                      > > Neumann, now a branch of Sennheiser (Oskar dummy Head and binaural mics for individuals like MKE2002) had their KU80, KU81 and KU100, the first with the mic inside the artificial ear canal, the latter with a mic close to the outer ear. This made recordings more versatile.
                      > > Here are pictures and a list:
                      > > http://www.green-sonic.net/deutsch/kunstkopf.html
                      > >
                      > > EMI central research lab developed their Sensaura recording technique based on artificial head mics and DSP-correction of the HRTF to make it playable via speakers. This created some coloration, but astounding soundstage with the RAF band marching your room nearly 90° on either side.
                      > >
                      > > I have done measurement and correction with TacT RCS, based on 2 mic positions, one for each ear. The effect is a very transparent soundstage.
                      > > There are lots of combfilter effects if one moves the head only slightly, which makes the system quite useless.
                      > >
                      > > Like Uli, I dont believe Staffelt.
                      > > Blauert (Spatial Hearing) and Theile have written a lot about this subject, some articles are available from the Neuman website.
                      > >
                      > > If the mic is placed at the ear, a headphone placed at the same position is best.
                      > > The signal travels the ear canal then once which is part of natural listening.
                      > > If the mic is placed inside the canal for a recording, the sound would pass the canal for a second time when listening.
                      > >
                      > > When listening with in-ear headphones, Siegfried Linkwitz has corrected the earcanal resonance because the canal is closed at either end this is relevant.
                      > > With the open ear the canal ends in a kind of horn shape and this IMO means coupling to the air without reflexion of the canals resonance frequency.
                      > > BR HM
                      > >
                      > >
                      > >
                      > > --- In regsaudioforum@yahoogroups.com, "Robert" <regonaudio@> wrote:
                      > > >
                      > > >
                      > > >
                      > > > I think we are somehow not getting the message of
                      > > > Staffelt's article, probably because I have not made
                      > > > it clear.
                      > > >
                      > > > According to his article, NO system that measures
                      > > > response in the abstract(in the absence of the head)
                      > > > is going to work right.
                      > > >
                      > > > His claim is that the thing that needs EQ(though NOT to flat)
                      > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                      > > > This can only be obtained by using dummy head measurement
                      > > > or even better measurement on your own head.
                      > > > Alternatively--and this it the REAL POINT--you should do this
                      > > > by listening.
                      > > >
                      > > > Almost everyone is ultimately disappointed in almost all the
                      > > > automatic systems, except the ones which have detailed user
                      > > > control.
                      > > >
                      > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                      > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                      > > >
                      > > > Engineers do not like this, philoisophically. It seems unscientific
                      > > > and "subjective" to them. But to me it makes perfect sense and
                      > > > also fits with my experience.
                      > > >
                      > > > I have really discovered that I do not like any automatic system
                      > > > for the finishing touches. ALL the usual models of how we hear sound in rooms are wrong, except the ones that are so complicated that one cannot use them or the ones that measure with your head in place.
                      > > > They have to be wrong, actually because of how complicated things are. A correct model would be a lot more complicated than a single frequency varying time window for example.
                      > > >
                      > > > But none of those latter kind(dummy head or your own head) are offered for sale.
                      > > > Every system on the market measures the sound in the absence of the head. And almost all of them try to deal with things by using time windows. According to Staffelt this won't work and in my experience it does not work, not exactly.
                      > > >
                      > > > None of this automatic stuff really works very well above the bass. But this is not to say that SOMETHING does not work well above the bass. As I used to say to GF, since you can change the sound in the higher frequencies you are almost bound to be able to make it better!
                      > > >
                      > > > The problem is the same as audio has always had. Engineers like engineering solutions, and solutions that involve human experience as an evaluation tool make engineers nervous, in general.
                      > > >
                      > > > I think it makes a whole lot of sense simply to get pink noise to sound right. Of course you have somehow to find out what right means, as TM suggested. But you can experiment and learn to hear what is right quite easily. Just try a lot of EQs both on pink noise and on recordings that you think ought to be good and set up something that makes them sound right in some intuitive sense.
                      > > >
                      > > > This could take some time! But it beats messing around trying to do the same thing by buying different equipment items or modifying your room. And it is a lot cheaper too.
                      > > >
                      > > > Of course one needs some sort of half decent listening room to begin with. But that is not so hard to arrange.
                      > > >
                      > > > I think the big failing here is human--people have not worked with user controlled EQ carefully enough. Talk about laziness! People would apparently rather buy stuff than work even a little.
                      > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                      > > > to play with EQ.
                      > > >
                      > > > I find it really intriguing to try to adjust one speaker to sound like another. Interesting to try!
                      > > >
                      > > > And don't forget that NO speaker, essentially , is as smooth as it ought to be. Just look at those NRC measurements. Forget questions of what radiation pattern ought to be for the moment and just look at one axis response. ALmost NO speaker is as smooth as it ought to be.
                      > > > It very seldom happens. And when adds in the effects of off axis--well, I think it is fair to say that it NEVER happens, that it is as smooth as it ought to be(independently of what "flat " means, everyone agrees that things ought to be really smooth--but they are not, not in detail. Even getting totally smooth 1/3 octave is quite unusual without EQ).
                      > > >
                      > > > Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).
                      > > >
                      > > > Engineers don't like this idea and most of them do not trust the customers either. But it is your music--and your head! whose transfer function you listen to.
                      > > >
                      > > > REG
                      > > >
                      > >
                      >
                    • Robert
                      Anyhting with complete and detailed user control is (except for questions of the phase changes associated to the EQ) the same as anything else, so this
                      Message 10 of 24 , Oct 6, 2010
                      • 0 Attachment
                        Anyhting with complete and detailed user control is
                        (except for questions of the phase changes associated to the EQ) the same as anything else, so this probably works ok
                        in that sense.

                        REG

                        --- In regsaudioforum@yahoogroups.com, "plinque_fish" <drplinque@...> wrote:
                        >
                        > It's worth looking at how Autocal works.
                        > It is automatic, but after initial calibration the user can make any further eq adjustments they want to via the computer, and then store those adjustments into each speaker so that in use the speakers do not have to be connected to an external PC. The software is calibrated (as far as I can tell) for the response of individual speakers and mics. I imagine it is would be very easy to eq the bass alone if that's all you wanted to do.
                        >
                        >
                        > > Anyway, anything with a name like Autocal is probably not really a total soultion above the bass --because if it is Auto it is going to produce almost surely the WRONG cal(ibration), the right one being a matter of listening.(I did not watch long enough to find out whether Autocal works above the bass--the name is just an example).
                        >
                      • Richard Tuck
                        * His claim is that the thing that needs EQ (though NOT to flat) is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL. This can only be obtained by
                        Message 11 of 24 , Oct 7, 2010
                        • 0 Attachment
                          * His claim is that the thing that needs EQ (though NOT to flat)
                          is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                          This can only be obtained by using dummy head measurement
                          or even better measurement on your own head.
                          Alternatively--and this it the REAL POINT--you should do this
                          by listening.*

                          I see a dilemma here, if I have got the right idea, "response AT THE
                          ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                          that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                          if you use just one speaker at a time?

                          Maybe one can use one of these recently developed hearing aids with a VERY
                          small microphone that is placed just outside the ear canal on a little
                          stalk.

                          Also with something like the Z-Systems is there a typical Q for the filters
                          doing the correction?

                          Richard

                          -----Original Message-----
                          From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                          On Behalf Of Robert
                          Sent: 06 October 2010 02:26
                          To: regsaudioforum@yahoogroups.com
                          Subject: [regsaudioforum] Listening and engineers



                          I think we are somehow not getting the message of
                          Staffelt's article, probably because I have not made
                          it clear.

                          According to his article, NO system that measures
                          response in the abstract(in the absence of the head)
                          is going to work right.

                          His claim is that the thing that needs EQ(though NOT to flat)
                          is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                          This can only be obtained by using dummy head measurement
                          or even better measurement on your own head.
                          Alternatively--and this it the REAL POINT--you should do this
                          by listening.

                          Almost everyone is ultimately disappointed in almost all the
                          automatic systems, except the ones which have detailed user
                          control.

                          YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                          Automatic roughs it out, but after that, you need to fine tune by ear.

                          Engineers do not like this, philoisophically. It seems unscientific
                          and "subjective" to them. But to me it makes perfect sense and
                          also fits with my experience.

                          I have really discovered that I do not like any automatic system
                          for the finishing touches. ALL the usual models of how we hear sound in
                          rooms are wrong, except the ones that are so complicated that one cannot use
                          them or the ones that measure with your head in place.
                          They have to be wrong, actually because of how complicated things are. A
                          correct model would be a lot more complicated than a single frequency
                          varying time window for example.

                          But none of those latter kind(dummy head or your own head) are offered for
                          sale.
                          Every system on the market measures the sound in the absence of the head.
                          And almost all of them try to deal with things by using time windows.
                          According to Staffelt this won't work and in my experience it does not work,
                          not exactly.

                          None of this automatic stuff really works very well above the bass. But this
                          is not to say that SOMETHING does not work well above the bass. As I used to
                          say to GF, since you can change the sound in the higher frequencies you are
                          almost bound to be able to make it better!

                          The problem is the same as audio has always had. Engineers like engineering
                          solutions, and solutions that involve human experience as an evaluation tool
                          make engineers nervous, in general.

                          I think it makes a whole lot of sense simply to get pink noise to sound
                          right. Of course you have somehow to find out what right means, as TM
                          suggested. But you can experiment and learn to hear what is right quite
                          easily. Just try a lot of EQs both on pink noise and on recordings that you
                          think ought to be good and set up something that makes them sound right in
                          some intuitive sense.

                          This could take some time! But it beats messing around trying to do the same
                          thing by buying different equipment items or modifying your room. And it is
                          a lot cheaper too.

                          Of course one needs some sort of half decent listening room to begin with.
                          But that is not so hard to arrange.

                          I think the big failing here is human--people have not worked with user
                          controlled EQ carefully enough. Talk about laziness! People would apparently
                          rather buy stuff than work even a little.
                          This is craziness to my mind. Anyway, once you get into it, it is FUN
                          to play with EQ.

                          I find it really intriguing to try to adjust one speaker to sound like
                          another. Interesting to try!

                          And don't forget that NO speaker, essentially , is as smooth as it ought to
                          be. Just look at those NRC measurements. Forget questions of what radiation
                          pattern ought to be for the moment and just look at one axis response.
                          ALmost NO speaker is as smooth as it ought to be.
                          It very seldom happens. And when adds in the effects of off axis--well, I
                          think it is fair to say that it NEVER happens, that it is as smooth as it
                          ought to be(independently of what "flat " means, everyone agrees that things
                          ought to be really smooth--but they are not, not in detail. Even getting
                          totally smooth 1/3 octave is quite unusual without EQ).

                          Anyway, anything with a name like Autocal is probably not really a total
                          soultion above the bass --because if it is Auto it is going to produce
                          almost surely the WRONG cal(ibration), the right one being a matter of
                          listening.(I did not watch long enough to find out whether Autocal works
                          above the bass--the name is just an example).

                          Engineers don't like this idea and most of them do not trust the customers
                          either. But it is your music--and your head! whose transfer function you
                          listen to.

                          REG




                          ------------------------------------

                          Yahoo! Groups Links
                        • Tom Mallin
                          From the current Zequalizer manual: Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4
                          Message 12 of 24 , Oct 7, 2010
                          • 0 Attachment
                            From the current Zequalizer manual:

                            "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."

                            The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.

                            On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.

                            >>> "Richard Tuck" <rtuck@...> 10/7/2010 8:19 AM >>>
                            * His claim is that the thing that needs EQ (though NOT to flat)
                            is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                            This can only be obtained by using dummy head measurement
                            or even better measurement on your own head.
                            Alternatively--and this it the REAL POINT--you should do this
                            by listening.*

                            I see a dilemma here, if I have got the right idea, "response AT THE
                            ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                            that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                            if you use just one speaker at a time?

                            Maybe one can use one of these recently developed hearing aids with a VERY
                            small microphone that is placed just outside the ear canal on a little
                            stalk.

                            Also with something like the Z-Systems is there a typical Q for the filters
                            doing the correction?

                            Richard

                            -----Original Message-----
                            From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                            On Behalf Of Robert
                            Sent: 06 October 2010 02:26
                            To: regsaudioforum@yahoogroups.com
                            Subject: [regsaudioforum] Listening and engineers



                            I think we are somehow not getting the message of
                            Staffelt's article, probably because I have not made
                            it clear.

                            According to his article, NO system that measures
                            response in the abstract(in the absence of the head)
                            is going to work right.

                            His claim is that the thing that needs EQ(though NOT to flat)
                            is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                            This can only be obtained by using dummy head measurement
                            or even better measurement on your own head.
                            Alternatively--and this it the REAL POINT--you should do this
                            by listening.

                            Almost everyone is ultimately disappointed in almost all the
                            automatic systems, except the ones which have detailed user
                            control.

                            YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                            Automatic roughs it out, but after that, you need to fine tune by ear.

                            Engineers do not like this, philoisophically. It seems unscientific
                            and "subjective" to them. But to me it makes perfect sense and
                            also fits with my experience.

                            I have really discovered that I do not like any automatic system
                            for the finishing touches. ALL the usual models of how we hear sound in
                            rooms are wrong, except the ones that are so complicated that one cannot use
                            them or the ones that measure with your head in place.
                            They have to be wrong, actually because of how complicated things are. A
                            correct model would be a lot more complicated than a single frequency
                            varying time window for example.

                            But none of those latter kind(dummy head or your own head) are offered for
                            sale.
                            Every system on the market measures the sound in the absence of the head.
                            And almost all of them try to deal with things by using time windows.
                            According to Staffelt this won't work and in my experience it does not work,
                            not exactly.

                            None of this automatic stuff really works very well above the bass. But this
                            is not to say that SOMETHING does not work well above the bass. As I used to
                            say to GF, since you can change the sound in the higher frequencies you are
                            almost bound to be able to make it better!

                            The problem is the same as audio has always had. Engineers like engineering
                            solutions, and solutions that involve human experience as an evaluation tool
                            make engineers nervous, in general.

                            I think it makes a whole lot of sense simply to get pink noise to sound
                            right. Of course you have somehow to find out what right means, as TM
                            suggested. But you can experiment and learn to hear what is right quite
                            easily. Just try a lot of EQs both on pink noise and on recordings that you
                            think ought to be good and set up something that makes them sound right in
                            some intuitive sense.

                            This could take some time! But it beats messing around trying to do the same
                            thing by buying different equipment items or modifying your room. And it is
                            a lot cheaper too.

                            Of course one needs some sort of half decent listening room to begin with.
                            But that is not so hard to arrange.

                            I think the big failing here is human--people have not worked with user
                            controlled EQ carefully enough. Talk about laziness! People would apparently
                            rather buy stuff than work even a little.
                            This is craziness to my mind. Anyway, once you get into it, it is FUN
                            to play with EQ.

                            I find it really intriguing to try to adjust one speaker to sound like
                            another. Interesting to try!

                            And don't forget that NO speaker, essentially , is as smooth as it ought to
                            be. Just look at those NRC measurements. Forget questions of what radiation
                            pattern ought to be for the moment and just look at one axis response.
                            ALmost NO speaker is as smooth as it ought to be.
                            It very seldom happens. And when adds in the effects of off axis--well, I
                            think it is fair to say that it NEVER happens, that it is as smooth as it
                            ought to be(independently of what "flat " means, everyone agrees that things
                            ought to be really smooth--but they are not, not in detail. Even getting
                            totally smooth 1/3 octave is quite unusual without EQ).

                            Anyway, anything with a name like Autocal is probably not really a total
                            soultion above the bass --because if it is Auto it is going to produce
                            almost surely the WRONG cal(ibration), the right one being a matter of
                            listening.(I did not watch long enough to find out whether Autocal works
                            above the bass--the name is just an example).

                            Engineers don't like this idea and most of them do not trust the customers
                            either. But it is your music--and your head! whose transfer function you
                            listen to.

                            REG




                            ------------------------------------

                            Yahoo! Groups Links





                            ------------------------------------

                            Yahoo! Groups Links
                          • uohh
                            With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                            Message 13 of 24 , Oct 7, 2010
                            • 0 Attachment
                              With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?

                              --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                              >
                              > From the current Zequalizer manual:
                              >
                              > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                              >
                              > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                              >
                              > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                              >
                              > >>> "Richard Tuck" <rtuck@...> 10/7/2010 8:19 AM >>>
                              > * His claim is that the thing that needs EQ (though NOT to flat)
                              > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                              > This can only be obtained by using dummy head measurement
                              > or even better measurement on your own head.
                              > Alternatively--and this it the REAL POINT--you should do this
                              > by listening.*
                              >
                              > I see a dilemma here, if I have got the right idea, "response AT THE
                              > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                              > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                              > if you use just one speaker at a time?
                              >
                              > Maybe one can use one of these recently developed hearing aids with a VERY
                              > small microphone that is placed just outside the ear canal on a little
                              > stalk.
                              >
                              > Also with something like the Z-Systems is there a typical Q for the filters
                              > doing the correction?
                              >
                              > Richard
                              >
                              > -----Original Message-----
                              > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                              > On Behalf Of Robert
                              > Sent: 06 October 2010 02:26
                              > To: regsaudioforum@yahoogroups.com
                              > Subject: [regsaudioforum] Listening and engineers
                              >
                              >
                              >
                              > I think we are somehow not getting the message of
                              > Staffelt's article, probably because I have not made
                              > it clear.
                              >
                              > According to his article, NO system that measures
                              > response in the abstract(in the absence of the head)
                              > is going to work right.
                              >
                              > His claim is that the thing that needs EQ(though NOT to flat)
                              > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                              > This can only be obtained by using dummy head measurement
                              > or even better measurement on your own head.
                              > Alternatively--and this it the REAL POINT--you should do this
                              > by listening.
                              >
                              > Almost everyone is ultimately disappointed in almost all the
                              > automatic systems, except the ones which have detailed user
                              > control.
                              >
                              > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                              > Automatic roughs it out, but after that, you need to fine tune by ear.
                              >
                              > Engineers do not like this, philoisophically. It seems unscientific
                              > and "subjective" to them. But to me it makes perfect sense and
                              > also fits with my experience.
                              >
                              > I have really discovered that I do not like any automatic system
                              > for the finishing touches. ALL the usual models of how we hear sound in
                              > rooms are wrong, except the ones that are so complicated that one cannot use
                              > them or the ones that measure with your head in place.
                              > They have to be wrong, actually because of how complicated things are. A
                              > correct model would be a lot more complicated than a single frequency
                              > varying time window for example.
                              >
                              > But none of those latter kind(dummy head or your own head) are offered for
                              > sale.
                              > Every system on the market measures the sound in the absence of the head.
                              > And almost all of them try to deal with things by using time windows.
                              > According to Staffelt this won't work and in my experience it does not work,
                              > not exactly.
                              >
                              > None of this automatic stuff really works very well above the bass. But this
                              > is not to say that SOMETHING does not work well above the bass. As I used to
                              > say to GF, since you can change the sound in the higher frequencies you are
                              > almost bound to be able to make it better!
                              >
                              > The problem is the same as audio has always had. Engineers like engineering
                              > solutions, and solutions that involve human experience as an evaluation tool
                              > make engineers nervous, in general.
                              >
                              > I think it makes a whole lot of sense simply to get pink noise to sound
                              > right. Of course you have somehow to find out what right means, as TM
                              > suggested. But you can experiment and learn to hear what is right quite
                              > easily. Just try a lot of EQs both on pink noise and on recordings that you
                              > think ought to be good and set up something that makes them sound right in
                              > some intuitive sense.
                              >
                              > This could take some time! But it beats messing around trying to do the same
                              > thing by buying different equipment items or modifying your room. And it is
                              > a lot cheaper too.
                              >
                              > Of course one needs some sort of half decent listening room to begin with.
                              > But that is not so hard to arrange.
                              >
                              > I think the big failing here is human--people have not worked with user
                              > controlled EQ carefully enough. Talk about laziness! People would apparently
                              > rather buy stuff than work even a little.
                              > This is craziness to my mind. Anyway, once you get into it, it is FUN
                              > to play with EQ.
                              >
                              > I find it really intriguing to try to adjust one speaker to sound like
                              > another. Interesting to try!
                              >
                              > And don't forget that NO speaker, essentially , is as smooth as it ought to
                              > be. Just look at those NRC measurements. Forget questions of what radiation
                              > pattern ought to be for the moment and just look at one axis response.
                              > ALmost NO speaker is as smooth as it ought to be.
                              > It very seldom happens. And when adds in the effects of off axis--well, I
                              > think it is fair to say that it NEVER happens, that it is as smooth as it
                              > ought to be(independently of what "flat " means, everyone agrees that things
                              > ought to be really smooth--but they are not, not in detail. Even getting
                              > totally smooth 1/3 octave is quite unusual without EQ).
                              >
                              > Anyway, anything with a name like Autocal is probably not really a total
                              > soultion above the bass --because if it is Auto it is going to produce
                              > almost surely the WRONG cal(ibration), the right one being a matter of
                              > listening.(I did not watch long enough to find out whether Autocal works
                              > above the bass--the name is just an example).
                              >
                              > Engineers don't like this idea and most of them do not trust the customers
                              > either. But it is your music--and your head! whose transfer function you
                              > listen to.
                              >
                              > REG
                              >
                              >
                              >
                              >
                              > ------------------------------------
                              >
                              > Yahoo! Groups Links
                              >
                              >
                              >
                              >
                              >
                              > ------------------------------------
                              >
                              > Yahoo! Groups Links
                              >
                            • Tom Mallin
                              I ve owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and
                              Message 14 of 24 , Oct 7, 2010
                              • 0 Attachment
                                I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT.  I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years. 
                                 
                                I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters.  Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                 
                                REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies.  Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening.  I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics.  With DSP, it's all just number crunching.

                                >>> "uohh" <uohh@...> 10/7/2010 3:24 PM >>>
                                With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?

                                --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                >
                                > From the current Zequalizer manual:
                                >
                                > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                >
                                > The Z-Systems EQ products have parametric equalizers with widely variable Q.  By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                >
                                > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments.  Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction.  With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose.  Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                >
                                > >>> "Richard Tuck" <rtuck@...> 10/7/2010 8:19 AM >>>
                                > * His claim is that the thing that needs EQ   (though NOT to flat)
                                > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                > This can only be obtained by using dummy head measurement
                                > or even better measurement on your own head.
                                > Alternatively--and this it the REAL POINT--you should do this
                                > by listening.*
                                >
                                > I see a dilemma here, if I have got the right idea, "response AT THE
                                > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                > that uses the HRTF to deal with inter-channel crosstalk.  Does it get easier
                                > if you use just one speaker at a time?
                                >
                                > Maybe one can use one of these recently developed hearing aids with a VERY
                                > small microphone that is placed just outside the ear canal on a little
                                > stalk.
                                >
                                > Also with something like the Z-Systems is there a typical Q for the filters
                                > doing the correction?
                                >
                                > Richard
                                >
                                > -----Original Message-----
                                > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                > On Behalf Of Robert
                                > Sent: 06 October 2010 02:26
                                > To: regsaudioforum@yahoogroups.com
                                > Subject: [regsaudioforum] Listening and engineers
                                >
                                >
                                >
                                > I think we are somehow not getting the message of
                                > Staffelt's article, probably because I have not made
                                > it clear.
                                >
                                > According to his article, NO system that measures
                                > response in the abstract(in the absence of the head)
                                > is going to work right.
                                >
                                > His claim is that the thing that needs EQ(though NOT to flat)
                                > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                > This can only be obtained by using dummy head measurement
                                > or even better measurement on your own head.
                                > Alternatively--and this it the REAL POINT--you should do this
                                > by listening.
                                >
                                > Almost everyone is ultimately disappointed in almost all the
                                > automatic systems, except the ones which have detailed user
                                > control.
                                >
                                > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                > Automatic roughs it out, but after that, you need to fine tune by ear.
                                >
                                > Engineers do not like this, philoisophically. It seems unscientific
                                > and "subjective" to them. But to me it makes perfect sense and
                                > also fits with my experience.
                                >
                                > I have really discovered that I do not like any automatic system
                                > for the finishing touches. ALL the usual models of how we hear sound in
                                > rooms are wrong, except the ones that are so complicated that one cannot use
                                > them or the ones that measure with your head in place.
                                > They have to be wrong, actually because of how complicated things are. A
                                > correct model would be a lot more complicated than a single frequency
                                > varying time window for example.
                                >
                                > But none of those latter kind(dummy head or your own head) are offered for
                                > sale.
                                > Every system on the market measures the sound in the absence of the head.
                                > And almost all of them try to deal with things by using time windows.
                                > According to Staffelt this won't work and in my experience it does not work,
                                > not exactly.
                                >
                                > None of this automatic stuff really works very well above the bass. But this
                                > is not to say that SOMETHING does not work well above the bass. As I used to
                                > say to GF, since you can change the sound in the higher frequencies you are
                                > almost bound to be able to make it better!
                                >
                                > The problem is the same as audio has always had. Engineers like engineering
                                > solutions, and solutions that involve human experience as an evaluation tool
                                > make engineers nervous, in general.
                                >
                                > I think it makes a whole lot of sense simply to get pink noise to sound
                                > right. Of course you have somehow to find out what right means, as TM
                                > suggested. But you can experiment and learn to hear what is right quite
                                > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                > think ought to be good and set up something that makes them sound right in
                                > some intuitive sense.
                                >
                                > This could take some time! But it beats messing around trying to do the same
                                > thing by buying different equipment items or modifying your room.  And it is
                                > a lot cheaper too.
                                >
                                > Of course one needs some sort of half decent listening room to begin with.
                                > But that is not so hard to arrange.
                                >
                                > I think the big failing here is human--people have not worked with user
                                > controlled EQ carefully enough. Talk about laziness! People would apparently
                                > rather buy stuff than work even a little.
                                > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                > to play with EQ.
                                >
                                > I find it really intriguing to try to adjust one speaker to sound like
                                > another. Interesting to try!
                                >
                                > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                > be. Just look at those NRC measurements. Forget questions of what radiation
                                > pattern ought to be for the moment and just look at one axis response.
                                > ALmost NO speaker is as smooth as it ought to be.
                                > It very seldom happens. And when adds in the effects of off axis--well, I
                                > think it is fair to say that it NEVER happens, that it is as smooth as it
                                > ought to be(independently of what "flat " means, everyone agrees that things
                                > ought to be really smooth--but they are not, not in detail. Even getting
                                > totally smooth 1/3 octave is quite unusual without EQ).
                                >
                                > Anyway, anything with a name like Autocal is probably not really a total
                                > soultion  above the bass --because if it is Auto it is going to produce
                                > almost surely the WRONG cal(ibration), the right one being a matter of
                                > listening.(I did not watch long enough to find out whether Autocal works
                                > above the bass--the name is just an example).
                                >
                                > Engineers don't like this idea and most of them do not trust the customers
                                > either. But it is your music--and your head! whose transfer function you
                                > listen to.
                                >
                                > REG
                                >
                                >
                                >
                                >
                                > ------------------------------------
                                >
                                > Yahoo! Groups Links
                                >
                                >
                                >
                                >
                                >
                                > ------------------------------------
                                >
                                > Yahoo! Groups Links
                                >




                                ------------------------------------

                                Yahoo! Groups Links

                                <*> To visit your group on the web, go to:
                                    http://groups.yahoo.com/group/regsaudioforum/

                                <*> Your email settings:
                                    Individual Email | Traditional

                                <*> To change settings online go to:
                                    http://groups.yahoo.com/group/regsaudioforum/join
                                    (Yahoo! ID required)

                                <*> To change settings via email:
                                    regsaudioforum-digest@yahoogroups.com
                                    regsaudioforum-fullfeatured@yahoogroups.com

                                <*> To unsubscribe from this group, send an email to:
                                    regsaudioforum-unsubscribe@yahoogroups.com

                                <*> Your use of Yahoo! Groups is subject to:
                                    http://docs.yahoo.com/info/terms/

                              • uohh
                                By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on
                                Message 15 of 24 , Oct 7, 2010
                                • 0 Attachment
                                  By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.

                                  And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.


                                  --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                  >
                                  > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT. I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years.
                                  >
                                  > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters. Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                  >
                                  > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies. Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening. I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics. With DSP, it's all just number crunching.
                                  >
                                  > >>> "uohh" <uohh@...> 10/7/2010 3:24 PM >>>
                                  > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                  >
                                  > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                  > >
                                  > > From the current Zequalizer manual:
                                  > >
                                  > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                  > >
                                  > > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                  > >
                                  > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                  > >
                                  > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                  > > * His claim is that the thing that needs EQ (though NOT to flat)
                                  > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                  > > This can only be obtained by using dummy head measurement
                                  > > or even better measurement on your own head.
                                  > > Alternatively--and this it the REAL POINT--you should do this
                                  > > by listening.*
                                  > >
                                  > > I see a dilemma here, if I have got the right idea, "response AT THE
                                  > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                  > > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                                  > > if you use just one speaker at a time?
                                  > >
                                  > > Maybe one can use one of these recently developed hearing aids with a VERY
                                  > > small microphone that is placed just outside the ear canal on a little
                                  > > stalk.
                                  > >
                                  > > Also with something like the Z-Systems is there a typical Q for the filters
                                  > > doing the correction?
                                  > >
                                  > > Richard
                                  > >
                                  > > -----Original Message-----
                                  > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                  > > On Behalf Of Robert
                                  > > Sent: 06 October 2010 02:26
                                  > > To: regsaudioforum@yahoogroups.com
                                  > > Subject: [regsaudioforum] Listening and engineers
                                  > >
                                  > >
                                  > >
                                  > > I think we are somehow not getting the message of
                                  > > Staffelt's article, probably because I have not made
                                  > > it clear.
                                  > >
                                  > > According to his article, NO system that measures
                                  > > response in the abstract(in the absence of the head)
                                  > > is going to work right.
                                  > >
                                  > > His claim is that the thing that needs EQ(though NOT to flat)
                                  > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                  > > This can only be obtained by using dummy head measurement
                                  > > or even better measurement on your own head.
                                  > > Alternatively--and this it the REAL POINT--you should do this
                                  > > by listening.
                                  > >
                                  > > Almost everyone is ultimately disappointed in almost all the
                                  > > automatic systems, except the ones which have detailed user
                                  > > control.
                                  > >
                                  > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                  > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                  > >
                                  > > Engineers do not like this, philoisophically. It seems unscientific
                                  > > and "subjective" to them. But to me it makes perfect sense and
                                  > > also fits with my experience.
                                  > >
                                  > > I have really discovered that I do not like any automatic system
                                  > > for the finishing touches. ALL the usual models of how we hear sound in
                                  > > rooms are wrong, except the ones that are so complicated that one cannot use
                                  > > them or the ones that measure with your head in place.
                                  > > They have to be wrong, actually because of how complicated things are. A
                                  > > correct model would be a lot more complicated than a single frequency
                                  > > varying time window for example.
                                  > >
                                  > > But none of those latter kind(dummy head or your own head) are offered for
                                  > > sale.
                                  > > Every system on the market measures the sound in the absence of the head.
                                  > > And almost all of them try to deal with things by using time windows.
                                  > > According to Staffelt this won't work and in my experience it does not work,
                                  > > not exactly.
                                  > >
                                  > > None of this automatic stuff really works very well above the bass. But this
                                  > > is not to say that SOMETHING does not work well above the bass. As I used to
                                  > > say to GF, since you can change the sound in the higher frequencies you are
                                  > > almost bound to be able to make it better!
                                  > >
                                  > > The problem is the same as audio has always had. Engineers like engineering
                                  > > solutions, and solutions that involve human experience as an evaluation tool
                                  > > make engineers nervous, in general.
                                  > >
                                  > > I think it makes a whole lot of sense simply to get pink noise to sound
                                  > > right. Of course you have somehow to find out what right means, as TM
                                  > > suggested. But you can experiment and learn to hear what is right quite
                                  > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                  > > think ought to be good and set up something that makes them sound right in
                                  > > some intuitive sense.
                                  > >
                                  > > This could take some time! But it beats messing around trying to do the same
                                  > > thing by buying different equipment items or modifying your room. And it is
                                  > > a lot cheaper too.
                                  > >
                                  > > Of course one needs some sort of half decent listening room to begin with.
                                  > > But that is not so hard to arrange.
                                  > >
                                  > > I think the big failing here is human--people have not worked with user
                                  > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                  > > rather buy stuff than work even a little.
                                  > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                  > > to play with EQ.
                                  > >
                                  > > I find it really intriguing to try to adjust one speaker to sound like
                                  > > another. Interesting to try!
                                  > >
                                  > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                  > > be. Just look at those NRC measurements. Forget questions of what radiation
                                  > > pattern ought to be for the moment and just look at one axis response.
                                  > > ALmost NO speaker is as smooth as it ought to be.
                                  > > It very seldom happens. And when adds in the effects of off axis--well, I
                                  > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                  > > ought to be(independently of what "flat " means, everyone agrees that things
                                  > > ought to be really smooth--but they are not, not in detail. Even getting
                                  > > totally smooth 1/3 octave is quite unusual without EQ).
                                  > >
                                  > > Anyway, anything with a name like Autocal is probably not really a total
                                  > > soultion above the bass --because if it is Auto it is going to produce
                                  > > almost surely the WRONG cal(ibration), the right one being a matter of
                                  > > listening.(I did not watch long enough to find out whether Autocal works
                                  > > above the bass--the name is just an example).
                                  > >
                                  > > Engineers don't like this idea and most of them do not trust the customers
                                  > > either. But it is your music--and your head! whose transfer function you
                                  > > listen to.
                                  > >
                                  > > REG
                                  > >
                                  > >
                                  > >
                                  > >
                                  > > ------------------------------------
                                  > >
                                  > > Yahoo! Groups Links
                                  > >
                                  > >
                                  > >
                                  > >
                                  > >
                                  > > ------------------------------------
                                  > >
                                  > > Yahoo! Groups Links
                                  > >
                                  >
                                  >
                                  >
                                  >
                                  > ------------------------------------
                                  >
                                  > Yahoo! Groups Links
                                  >
                                • Tom Mallin
                                  My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than
                                  Message 16 of 24 , Oct 7, 2010
                                  • 0 Attachment
                                    My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000.  So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                     
                                    Yes, near-field listening reduces room effects heard at the listening position as to high frequencies.  But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s.  While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well.  In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up.  Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                     
                                    FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters.  It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency.  Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.

                                    >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                    By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.

                                    And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.


                                    --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                    >
                                    > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT.  I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years. 

                                    > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters.  Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.

                                    > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies.  Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening.  I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics.  With DSP, it's all just number crunching.
                                    >
                                    > >>> "uohh" <uohh@...> 10/7/2010 3:24 PM >>>
                                    > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                    >
                                    > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                    > >
                                    > > From the current Zequalizer manual:
                                    > >
                                    > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                    > >
                                    > > The Z-Systems EQ products have parametric equalizers with widely variable Q.  By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                    > >
                                    > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments.  Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction.  With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose.  Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                    > >
                                    > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                    > > * His claim is that the thing that needs EQ   (though NOT to flat)
                                    > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                    > > This can only be obtained by using dummy head measurement
                                    > > or even better measurement on your own head.
                                    > > Alternatively--and this it the REAL POINT--you should do this
                                    > > by listening.*
                                    > >
                                    > > I see a dilemma here, if I have got the right idea, "response AT THE
                                    > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                    > > that uses the HRTF to deal with inter-channel crosstalk.  Does it get easier
                                    > > if you use just one speaker at a time?
                                    > >
                                    > > Maybe one can use one of these recently developed hearing aids with a VERY
                                    > > small microphone that is placed just outside the ear canal on a little
                                    > > stalk.
                                    > >
                                    > > Also with something like the Z-Systems is there a typical Q for the filters
                                    > > doing the correction?
                                    > >
                                    > > Richard
                                    > >
                                    > > -----Original Message-----
                                    > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                    > > On Behalf Of Robert
                                    > > Sent: 06 October 2010 02:26
                                    > > To: regsaudioforum@yahoogroups.com
                                    > > Subject: [regsaudioforum] Listening and engineers
                                    > >
                                    > >
                                    > >
                                    > > I think we are somehow not getting the message of
                                    > > Staffelt's article, probably because I have not made
                                    > > it clear.
                                    > >
                                    > > According to his article, NO system that measures
                                    > > response in the abstract(in the absence of the head)
                                    > > is going to work right.
                                    > >
                                    > > His claim is that the thing that needs EQ(though NOT to flat)
                                    > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                    > > This can only be obtained by using dummy head measurement
                                    > > or even better measurement on your own head.
                                    > > Alternatively--and this it the REAL POINT--you should do this
                                    > > by listening.
                                    > >
                                    > > Almost everyone is ultimately disappointed in almost all the
                                    > > automatic systems, except the ones which have detailed user
                                    > > control.
                                    > >
                                    > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                    > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                    > >
                                    > > Engineers do not like this, philoisophically. It seems unscientific
                                    > > and "subjective" to them. But to me it makes perfect sense and
                                    > > also fits with my experience.
                                    > >
                                    > > I have really discovered that I do not like any automatic system
                                    > > for the finishing touches. ALL the usual models of how we hear sound in
                                    > > rooms are wrong, except the ones that are so complicated that one cannot use
                                    > > them or the ones that measure with your head in place.
                                    > > They have to be wrong, actually because of how complicated things are. A
                                    > > correct model would be a lot more complicated than a single frequency
                                    > > varying time window for example.
                                    > >
                                    > > But none of those latter kind(dummy head or your own head) are offered for
                                    > > sale.
                                    > > Every system on the market measures the sound in the absence of the head.
                                    > > And almost all of them try to deal with things by using time windows.
                                    > > According to Staffelt this won't work and in my experience it does not work,
                                    > > not exactly.
                                    > >
                                    > > None of this automatic stuff really works very well above the bass. But this
                                    > > is not to say that SOMETHING does not work well above the bass. As I used to
                                    > > say to GF, since you can change the sound in the higher frequencies you are
                                    > > almost bound to be able to make it better!
                                    > >
                                    > > The problem is the same as audio has always had. Engineers like engineering
                                    > > solutions, and solutions that involve human experience as an evaluation tool
                                    > > make engineers nervous, in general.
                                    > >
                                    > > I think it makes a whole lot of sense simply to get pink noise to sound
                                    > > right. Of course you have somehow to find out what right means, as TM
                                    > > suggested. But you can experiment and learn to hear what is right quite
                                    > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                    > > think ought to be good and set up something that makes them sound right in
                                    > > some intuitive sense.
                                    > >
                                    > > This could take some time! But it beats messing around trying to do the same
                                    > > thing by buying different equipment items or modifying your room.  And it is
                                    > > a lot cheaper too.
                                    > >
                                    > > Of course one needs some sort of half decent listening room to begin with.
                                    > > But that is not so hard to arrange.
                                    > >
                                    > > I think the big failing here is human--people have not worked with user
                                    > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                    > > rather buy stuff than work even a little.
                                    > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                    > > to play with EQ.
                                    > >
                                    > > I find it really intriguing to try to adjust one speaker to sound like
                                    > > another. Interesting to try!
                                    > >
                                    > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                    > > be. Just look at those NRC measurements. Forget questions of what radiation
                                    > > pattern ought to be for the moment and just look at one axis response.
                                    > > ALmost NO speaker is as smooth as it ought to be.
                                    > > It very seldom happens. And when adds in the effects of off axis--well, I
                                    > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                    > > ought to be(independently of what "flat " means, everyone agrees that things
                                    > > ought to be really smooth--but they are not, not in detail. Even getting
                                    > > totally smooth 1/3 octave is quite unusual without EQ).
                                    > >
                                    > > Anyway, anything with a name like Autocal is probably not really a total
                                    > > soultion  above the bass --because if it is Auto it is going to produce
                                    > > almost surely the WRONG cal(ibration), the right one being a matter of
                                    > > listening.(I did not watch long enough to find out whether Autocal works
                                    > > above the bass--the name is just an example).
                                    > >
                                    > > Engineers don't like this idea and most of them do not trust the customers
                                    > > either. But it is your music--and your head! whose transfer function you
                                    > > listen to.
                                    > >
                                    > > REG
                                    > >
                                    > >
                                    > >
                                    > >
                                    > > ------------------------------------
                                    > >
                                    > > Yahoo! Groups Links
                                    > >
                                    > >
                                    > >
                                    > >
                                    > >
                                    > > ------------------------------------
                                    > >
                                    > > Yahoo! Groups Links
                                    > >
                                    >
                                    >
                                    >
                                    >
                                    > ------------------------------------
                                    >
                                    > Yahoo! Groups Links
                                    >




                                    ------------------------------------

                                    Yahoo! Groups Links

                                    <*> To visit your group on the web, go to:
                                        http://groups.yahoo.com/group/regsaudioforum/

                                    <*> Your email settings:
                                        Individual Email | Traditional

                                    <*> To change settings online go to:
                                        http://groups.yahoo.com/group/regsaudioforum/join
                                        (Yahoo! ID required)

                                    <*> To change settings via email:
                                        regsaudioforum-digest@yahoogroups.com
                                        regsaudioforum-fullfeatured@yahoogroups.com

                                    <*> To unsubscribe from this group, send an email to:
                                        regsaudioforum-unsubscribe@yahoogroups.com

                                    <*> Your use of Yahoo! Groups is subject to:
                                        http://docs.yahoo.com/info/terms/

                                  • uohh
                                    Hi Tom, I only have experience with a Tact 2s and a 2.2x myself, I m trying to understand why you would choose the parametric equalizer function over the use
                                    Message 17 of 24 , Oct 8, 2010
                                    • 0 Attachment
                                      Hi Tom,

                                      I only have experience with a Tact 2s and a 2.2x myself, I'm trying to understand why you would choose the parametric equalizer function over the use of plotting your own target curves, whether there is an audible benefit doing so.

                                      FYI, we 2.2x model users have a way to bypass correction of any frequency range by independently tracing the left and right target curves to that of the measurement curves.

                                      Joseph


                                      --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                      >
                                      > My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000. So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                      >
                                      > Yes, near-field listening reduces room effects heard at the listening position as to high frequencies. But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s. While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well. In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up. Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                      >
                                      > FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters. It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency. Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.
                                      >
                                      > >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                      > By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.
                                      >
                                      > And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.
                                      >
                                      >
                                      > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                      > >
                                      > > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT. I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years.
                                      > >
                                      > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters. Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                      > >
                                      > > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies. Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening. I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics. With DSP, it's all just number crunching.
                                      > >
                                      > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                      > > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                      > >
                                      > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                      > > >
                                      > > > From the current Zequalizer manual:
                                      > > >
                                      > > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                      > > >
                                      > > > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                      > > >
                                      > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                      > > >
                                      > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                      > > > * His claim is that the thing that needs EQ (though NOT to flat)
                                      > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                      > > > This can only be obtained by using dummy head measurement
                                      > > > or even better measurement on your own head.
                                      > > > Alternatively--and this it the REAL POINT--you should do this
                                      > > > by listening.*
                                      > > >
                                      > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                      > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                      > > > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                                      > > > if you use just one speaker at a time?
                                      > > >
                                      > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                      > > > small microphone that is placed just outside the ear canal on a little
                                      > > > stalk.
                                      > > >
                                      > > > Also with something like the Z-Systems is there a typical Q for the filters
                                      > > > doing the correction?
                                      > > >
                                      > > > Richard
                                      > > >
                                      > > > -----Original Message-----
                                      > > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                      > > > On Behalf Of Robert
                                      > > > Sent: 06 October 2010 02:26
                                      > > > To: regsaudioforum@yahoogroups.com
                                      > > > Subject: [regsaudioforum] Listening and engineers
                                      > > >
                                      > > >
                                      > > >
                                      > > > I think we are somehow not getting the message of
                                      > > > Staffelt's article, probably because I have not made
                                      > > > it clear.
                                      > > >
                                      > > > According to his article, NO system that measures
                                      > > > response in the abstract(in the absence of the head)
                                      > > > is going to work right.
                                      > > >
                                      > > > His claim is that the thing that needs EQ(though NOT to flat)
                                      > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                      > > > This can only be obtained by using dummy head measurement
                                      > > > or even better measurement on your own head.
                                      > > > Alternatively--and this it the REAL POINT--you should do this
                                      > > > by listening.
                                      > > >
                                      > > > Almost everyone is ultimately disappointed in almost all the
                                      > > > automatic systems, except the ones which have detailed user
                                      > > > control.
                                      > > >
                                      > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                      > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                      > > >
                                      > > > Engineers do not like this, philoisophically. It seems unscientific
                                      > > > and "subjective" to them. But to me it makes perfect sense and
                                      > > > also fits with my experience.
                                      > > >
                                      > > > I have really discovered that I do not like any automatic system
                                      > > > for the finishing touches. ALL the usual models of how we hear sound in
                                      > > > rooms are wrong, except the ones that are so complicated that one cannot use
                                      > > > them or the ones that measure with your head in place.
                                      > > > They have to be wrong, actually because of how complicated things are. A
                                      > > > correct model would be a lot more complicated than a single frequency
                                      > > > varying time window for example.
                                      > > >
                                      > > > But none of those latter kind(dummy head or your own head) are offered for
                                      > > > sale.
                                      > > > Every system on the market measures the sound in the absence of the head.
                                      > > > And almost all of them try to deal with things by using time windows.
                                      > > > According to Staffelt this won't work and in my experience it does not work,
                                      > > > not exactly.
                                      > > >
                                      > > > None of this automatic stuff really works very well above the bass. But this
                                      > > > is not to say that SOMETHING does not work well above the bass. As I used to
                                      > > > say to GF, since you can change the sound in the higher frequencies you are
                                      > > > almost bound to be able to make it better!
                                      > > >
                                      > > > The problem is the same as audio has always had. Engineers like engineering
                                      > > > solutions, and solutions that involve human experience as an evaluation tool
                                      > > > make engineers nervous, in general.
                                      > > >
                                      > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                      > > > right. Of course you have somehow to find out what right means, as TM
                                      > > > suggested. But you can experiment and learn to hear what is right quite
                                      > > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                      > > > think ought to be good and set up something that makes them sound right in
                                      > > > some intuitive sense.
                                      > > >
                                      > > > This could take some time! But it beats messing around trying to do the same
                                      > > > thing by buying different equipment items or modifying your room. And it is
                                      > > > a lot cheaper too.
                                      > > >
                                      > > > Of course one needs some sort of half decent listening room to begin with.
                                      > > > But that is not so hard to arrange.
                                      > > >
                                      > > > I think the big failing here is human--people have not worked with user
                                      > > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                      > > > rather buy stuff than work even a little.
                                      > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                      > > > to play with EQ.
                                      > > >
                                      > > > I find it really intriguing to try to adjust one speaker to sound like
                                      > > > another. Interesting to try!
                                      > > >
                                      > > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                      > > > be. Just look at those NRC measurements. Forget questions of what radiation
                                      > > > pattern ought to be for the moment and just look at one axis response.
                                      > > > ALmost NO speaker is as smooth as it ought to be.
                                      > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                      > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                      > > > ought to be(independently of what "flat " means, everyone agrees that things
                                      > > > ought to be really smooth--but they are not, not in detail. Even getting
                                      > > > totally smooth 1/3 octave is quite unusual without EQ).
                                      > > >
                                      > > > Anyway, anything with a name like Autocal is probably not really a total
                                      > > > soultion above the bass --because if it is Auto it is going to produce
                                      > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                      > > > listening.(I did not watch long enough to find out whether Autocal works
                                      > > > above the bass--the name is just an example).
                                      > > >
                                      > > > Engineers don't like this idea and most of them do not trust the customers
                                      > > > either. But it is your music--and your head! whose transfer function you
                                      > > > listen to.
                                      > > >
                                      > > > REG
                                      > > >
                                      > > >
                                      > > >
                                      > > >
                                      > > > ------------------------------------
                                      > > >
                                      > > > Yahoo! Groups Links
                                      > > >
                                      > > >
                                      > > >
                                      > > >
                                      > > >
                                      > > > ------------------------------------
                                      > > >
                                      > > > Yahoo! Groups Links
                                      > > >
                                      > >
                                      > >
                                      > >
                                      > >
                                      > > ------------------------------------
                                      > >
                                      > > Yahoo! Groups Links
                                      > >
                                      >
                                      >
                                      >
                                      >
                                      > ------------------------------------
                                      >
                                      > Yahoo! Groups Links
                                      >
                                    • Tom Mallin
                                      I am just getting back to using the EQ functions of the TacT after about a year of not doing so. I ll have to try it and see which sounds better: target
                                      Message 18 of 24 , Oct 8, 2010
                                      • 0 Attachment
                                        I am just getting back to using the EQ functions of the TacT after about a year of not doing so.  I'll have to try it and see which sounds better:  target curve or parametric.  You can do the same.  That is kind of the point of REG's recent comments about Staffelt's research:  the ear is the final arbiter and perhaps adjustment as broad brush as 1/3-octave bands is the way to go.

                                        >>> "uohh" <uohh@...> 10/8/2010 4:10 AM >>>
                                        Hi Tom,

                                        I only have experience with a Tact 2s and a 2.2x myself, I'm trying to understand why you would choose the parametric equalizer function over the use of plotting your own target curves, whether there is an audible benefit doing so.

                                        FYI, we 2.2x model users have a way to bypass correction of any frequency range by independently tracing the left and right target curves to that of the measurement curves.

                                        Joseph


                                        --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                        >
                                        > My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000.  So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.

                                        > Yes, near-field listening reduces room effects heard at the listening position as to high frequencies.  But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s.  While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well.  In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up.  Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.

                                        > FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters.  It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency.  Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.
                                        >
                                        > >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                        > By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.
                                        >
                                        > And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.
                                        >
                                        >
                                        > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                        > >
                                        > > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT.  I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years. 
                                        > > 
                                        > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters.  Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                        > > 
                                        > > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies.  Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening.  I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics.  With DSP, it's all just number crunching.
                                        > >
                                        > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                        > > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                        > >
                                        > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                        > > >
                                        > > > From the current Zequalizer manual:
                                        > > >
                                        > > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                        > > >
                                        > > > The Z-Systems EQ products have parametric equalizers with widely variable Q.  By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                        > > >
                                        > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments.  Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction.  With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose.  Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                        > > >
                                        > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                        > > > * His claim is that the thing that needs EQ   (though NOT to flat)
                                        > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                        > > > This can only be obtained by using dummy head measurement
                                        > > > or even better measurement on your own head.
                                        > > > Alternatively--and this it the REAL POINT--you should do this
                                        > > > by listening.*
                                        > > >
                                        > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                        > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                        > > > that uses the HRTF to deal with inter-channel crosstalk.  Does it get easier
                                        > > > if you use just one speaker at a time?
                                        > > >
                                        > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                        > > > small microphone that is placed just outside the ear canal on a little
                                        > > > stalk.
                                        > > >
                                        > > > Also with something like the Z-Systems is there a typical Q for the filters
                                        > > > doing the correction?
                                        > > >
                                        > > > Richard
                                        > > >
                                        > > > -----Original Message-----
                                        > > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                        > > > On Behalf Of Robert
                                        > > > Sent: 06 October 2010 02:26
                                        > > > To: regsaudioforum@yahoogroups.com
                                        > > > Subject: [regsaudioforum] Listening and engineers
                                        > > >
                                        > > >
                                        > > >
                                        > > > I think we are somehow not getting the message of
                                        > > > Staffelt's article, probably because I have not made
                                        > > > it clear.
                                        > > >
                                        > > > According to his article, NO system that measures
                                        > > > response in the abstract(in the absence of the head)
                                        > > > is going to work right.
                                        > > >
                                        > > > His claim is that the thing that needs EQ(though NOT to flat)
                                        > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                        > > > This can only be obtained by using dummy head measurement
                                        > > > or even better measurement on your own head.
                                        > > > Alternatively--and this it the REAL POINT--you should do this
                                        > > > by listening.
                                        > > >
                                        > > > Almost everyone is ultimately disappointed in almost all the
                                        > > > automatic systems, except the ones which have detailed user
                                        > > > control.
                                        > > >
                                        > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                        > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                        > > >
                                        > > > Engineers do not like this, philoisophically. It seems unscientific
                                        > > > and "subjective" to them. But to me it makes perfect sense and
                                        > > > also fits with my experience.
                                        > > >
                                        > > > I have really discovered that I do not like any automatic system
                                        > > > for the finishing touches. ALL the usual models of how we hear sound in
                                        > > > rooms are wrong, except the ones that are so complicated that one cannot use
                                        > > > them or the ones that measure with your head in place.
                                        > > > They have to be wrong, actually because of how complicated things are. A
                                        > > > correct model would be a lot more complicated than a single frequency
                                        > > > varying time window for example.
                                        > > >
                                        > > > But none of those latter kind(dummy head or your own head) are offered for
                                        > > > sale.
                                        > > > Every system on the market measures the sound in the absence of the head.
                                        > > > And almost all of them try to deal with things by using time windows.
                                        > > > According to Staffelt this won't work and in my experience it does not work,
                                        > > > not exactly.
                                        > > >
                                        > > > None of this automatic stuff really works very well above the bass. But this
                                        > > > is not to say that SOMETHING does not work well above the bass. As I used to
                                        > > > say to GF, since you can change the sound in the higher frequencies you are
                                        > > > almost bound to be able to make it better!
                                        > > >
                                        > > > The problem is the same as audio has always had. Engineers like engineering
                                        > > > solutions, and solutions that involve human experience as an evaluation tool
                                        > > > make engineers nervous, in general.
                                        > > >
                                        > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                        > > > right. Of course you have somehow to find out what right means, as TM
                                        > > > suggested. But you can experiment and learn to hear what is right quite
                                        > > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                        > > > think ought to be good and set up something that makes them sound right in
                                        > > > some intuitive sense.
                                        > > >
                                        > > > This could take some time! But it beats messing around trying to do the same
                                        > > > thing by buying different equipment items or modifying your room.  And it is
                                        > > > a lot cheaper too.
                                        > > >
                                        > > > Of course one needs some sort of half decent listening room to begin with.
                                        > > > But that is not so hard to arrange.
                                        > > >
                                        > > > I think the big failing here is human--people have not worked with user
                                        > > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                        > > > rather buy stuff than work even a little.
                                        > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                        > > > to play with EQ.
                                        > > >
                                        > > > I find it really intriguing to try to adjust one speaker to sound like
                                        > > > another. Interesting to try!
                                        > > >
                                        > > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                        > > > be. Just look at those NRC measurements. Forget questions of what radiation
                                        > > > pattern ought to be for the moment and just look at one axis response.
                                        > > > ALmost NO speaker is as smooth as it ought to be.
                                        > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                        > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                        > > > ought to be(independently of what "flat " means, everyone agrees that things
                                        > > > ought to be really smooth--but they are not, not in detail. Even getting
                                        > > > totally smooth 1/3 octave is quite unusual without EQ).
                                        > > >
                                        > > > Anyway, anything with a name like Autocal is probably not really a total
                                        > > > soultion  above the bass --because if it is Auto it is going to produce
                                        > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                        > > > listening.(I did not watch long enough to find out whether Autocal works
                                        > > > above the bass--the name is just an example).
                                        > > >
                                        > > > Engineers don't like this idea and most of them do not trust the customers
                                        > > > either. But it is your music--and your head! whose transfer function you
                                        > > > listen to.
                                        > > >
                                        > > > REG
                                        > > >
                                        > > >
                                        > > >
                                        > > >
                                        > > > ------------------------------------
                                        > > >
                                        > > > Yahoo! Groups Links
                                        > > >
                                        > > >
                                        > > >
                                        > > >
                                        > > >
                                        > > > ------------------------------------
                                        > > >
                                        > > > Yahoo! Groups Links
                                        > > >
                                        > >
                                        > >
                                        > >
                                        > >
                                        > > ------------------------------------
                                        > >
                                        > > Yahoo! Groups Links
                                        > >
                                        >
                                        >
                                        >
                                        >
                                        > ------------------------------------
                                        >
                                        > Yahoo! Groups Links
                                        >




                                        ------------------------------------

                                        Yahoo! Groups Links

                                        <*> To visit your group on the web, go to:
                                            http://groups.yahoo.com/group/regsaudioforum/

                                        <*> Your email settings:
                                            Individual Email | Traditional

                                        <*> To change settings online go to:
                                            http://groups.yahoo.com/group/regsaudioforum/join
                                            (Yahoo! ID required)

                                        <*> To change settings via email:
                                            regsaudioforum-digest@yahoogroups.com
                                            regsaudioforum-fullfeatured@yahoogroups.com

                                        <*> To unsubscribe from this group, send an email to:
                                            regsaudioforum-unsubscribe@yahoogroups.com

                                        <*> Your use of Yahoo! Groups is subject to:
                                            http://docs.yahoo.com/info/terms/

                                      • ymm
                                        ________________________________ From: Tom Mallin To: regsaudioforum@yahoogroups.com Sent: Friday, 8 October 2010 12:44:48 Subject:
                                        Message 19 of 24 , Oct 8, 2010
                                        • 0 Attachment



                                          From: Tom Mallin <tmallin@...>
                                          To: regsaudioforum@yahoogroups.com
                                          Sent: Friday, 8 October 2010 12:44:48
                                          Subject: [regsaudioforum] Re: Listening and engineers

                                           

                                          My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000.  So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                           
                                          Yes, near-field listening reduces room effects heard at the listening position as to high frequencies.  But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s.  While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well.  In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up.  Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                           
                                          FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters.  It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency.  Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.

                                          >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                          By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.

                                          And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.


                                          --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                          >
                                          > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT.  I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years. 

                                          > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters.  Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.

                                          > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies.  Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening.  I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics.  With DSP, it's all just number crunching.
                                          >
                                          > >>> "uohh" <uohh@...> 10/7/2010 3:24 PM >>>
                                          > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                          >
                                          > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                          > >
                                          > > From the current Zequalizer manual:
                                          > >
                                          > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                          > >
                                          > > The Z-Systems EQ products have parametric equalizers with widely variable Q.  By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                          > >
                                          > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments.  Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction.  With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose.  Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                          > >
                                          > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                          > > * His claim is that the thing that needs EQ   (though NOT to flat)
                                          > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                          > > This can only be obtained by using dummy head measurement
                                          > > or even better measurement on your own head.
                                          > > Alternatively--and this it the REAL POINT--you should do this
                                          > > by listening.*
                                          > >
                                          > > I see a dilemma here, if I have got the right idea, "response AT THE
                                          > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                          > > that uses the HRTF to deal with inter-channel crosstalk.  Does it get easier
                                          > > if you use just one speaker at a time?
                                          > >
                                          > > Maybe one can use one of these recently developed hearing aids with a VERY
                                          > > small microphone that is placed just outside the ear canal on a little
                                          > > stalk.
                                          > >
                                          > > Also with something like the Z-Systems is there a typical Q for the filters
                                          > > doing the correction?
                                          > >
                                          > > Richard
                                          > >
                                          > > -----Original Message-----
                                          > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                          > > On Behalf Of Robert
                                          > > Sent: 06 October 2010 02:26
                                          > > To: regsaudioforum@yahoogroups.com
                                          > > Subject: [regsaudioforum] Listening and engineers
                                          > >
                                          > >
                                          > >
                                          > > I think we are somehow not getting the message of
                                          > > Staffelt's article, probably because I have not made
                                          > > it clear.
                                          > >
                                          > > According to his article, NO system that measures
                                          > > response in the abstract(in the absence of the head)
                                          > > is going to work right.
                                          > >
                                          > > His claim is that the thing that needs EQ(though NOT to flat)
                                          > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                          > > This can only be obtained by using dummy head measurement
                                          > > or even better measurement on your own head.
                                          > > Alternatively--and this it the REAL POINT--you should do this
                                          > > by listening.
                                          > >
                                          > > Almost everyone is ultimately disappointed in almost all the
                                          > > automatic systems, except the ones which have detailed user
                                          > > control.
                                          > >
                                          > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                          > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                          > >
                                          > > Engineers do not like this, philoisophically. It seems unscientific
                                          > > and "subjective" to them. But to me it makes perfect sense and
                                          > > also fits with my experience.
                                          > >
                                          > > I have really discovered that I do not like any automatic system
                                          > > for the finishing touches. ALL the usual models of how we hear sound in
                                          > > rooms are wrong, except the ones that are so complicated that one cannot use
                                          > > them or the ones that measure with your head in place.
                                          > > They have to be wrong, actually because of how complicated things are. A
                                          > > correct model would be a lot more complicated than a single frequency
                                          > > varying time window for example.
                                          > >
                                          > > But none of those latter kind(dummy head or your own head) are offered for
                                          > > sale.
                                          > > Every system on the market measures the sound in the absence of the head.
                                          > > And almost all of them try to deal with things by using time windows.
                                          > > According to Staffelt this won't work and in my experience it does not work,
                                          > > not exactly.
                                          > >
                                          > > None of this automatic stuff really works very well above the bass. But this
                                          > > is not to say that SOMETHING does not work well above the bass. As I used to
                                          > > say to GF, since you can change the sound in the higher frequencies you are
                                          > > almost bound to be able to make it better!
                                          > >
                                          > > The problem is the same as audio has always had. Engineers like engineering
                                          > > solutions, and solutions that involve human experience as an evaluation tool
                                          > > make engineers nervous, in general.
                                          > >
                                          > > I think it makes a whole lot of sense simply to get pink noise to sound
                                          > > right. Of course you have somehow to find out what right means, as TM
                                          > > suggested. But you can experiment and learn to hear what is right quite
                                          > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                          > > think ought to be good and set up something that makes them sound right in
                                          > > some intuitive sense.
                                          > >
                                          > > This could take some time! But it beats messing around trying to do the same
                                          > > thing by buying different equipment items or modifying your room.  And it is
                                          > > a lot cheaper too.
                                          > >
                                          > > Of course one needs some sort of half decent listening room to begin with.
                                          > > But that is not so hard to arrange.
                                          > >
                                          > > I think the big failing here is human--people have not worked with user
                                          > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                          > > rather buy stuff than work even a little.
                                          > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                          > > to play with EQ.
                                          > >
                                          > > I find it really intriguing to try to adjust one speaker to sound like
                                          > > another. Interesting to try!
                                          > >
                                          > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                          > > be. Just look at those NRC measurements. Forget questions of what radiation
                                          > > pattern ought to be for the moment and just look at one axis response.
                                          > > ALmost NO speaker is as smooth as it ought to be.
                                          > > It very seldom happens. And when adds in the effects of off axis--well, I
                                          > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                          > > ought to be(independently of what "flat " means, everyone agrees that things
                                          > > ought to be really smooth--but they are not, not in detail. Even getting
                                          > > totally smooth 1/3 octave is quite unusual without EQ).
                                          > >
                                          > > Anyway, anything with a name like Autocal is probably not really a total
                                          > > soultion  above the bass --because if it is Auto it is going to produce
                                          > > almost surely the WRONG cal(ibration), the right one being a matter of
                                          > > listening.(I did not watch long enough to find out whether Autocal works
                                          > > above the bass--the name is just an example).
                                          > >
                                          > > Engineers don't like this idea and most of them do not trust the customers
                                          > > either. But it is your music--and your head! whose transfer function you
                                          > > listen to.
                                          > >
                                          > > REG
                                          > >
                                          > >
                                          > >
                                          > >
                                          > > ------------------------------------
                                          > >
                                          > > Yahoo! Groups Links
                                          > >
                                          > >
                                          > >
                                          > >
                                          > >
                                          > > ------------------------------------
                                          > >
                                          > > Yahoo! Groups Links
                                          > >
                                          >
                                          >
                                          >
                                          >
                                          > ------------------------------------
                                          >
                                          > Yahoo! Groups Links
                                          >




                                          ------------------------------------

                                          Yahoo! Groups Links

                                          <*> To visit your group on the web, go to:
                                              http://groups.yahoo.com/group/regsaudioforum/

                                          <*> Your email settings:
                                              Individual Email | Traditional

                                          <*> To change settings online go to:
                                              http://groups.yahoo.com/group/regsaudioforum/join
                                              (Yahoo! ID required)

                                          <*> To change settings via email:
                                              regsaudioforum-digest@yahoogroups.com
                                              regsaudioforum-fullfeatured@yahoogroups.com

                                          <*> To unsubscribe from this group, send an email to:
                                              regsaudioforum-unsubscribe@yahoogroups.com

                                          <*> Your use of Yahoo! Groups is subject to:
                                              http://docs.yahoo.com/info/terms/


                                        • uohh
                                          Would love to hear your findings!
                                          Message 20 of 24 , Oct 8, 2010
                                          • 0 Attachment
                                            Would love to hear your findings!

                                            --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                            >
                                            > I am just getting back to using the EQ functions of the TacT after about a year of not doing so. I'll have to try it and see which sounds better: target curve or parametric. You can do the same. That is kind of the point of REG's recent comments about Staffelt's research: the ear is the final arbiter and perhaps adjustment as broad brush as 1/3-octave bands is the way to go.
                                            >
                                            > >>> "uohh" <uohh@...> 10/8/2010 4:10 AM >>>
                                            > Hi Tom,
                                            >
                                            > I only have experience with a Tact 2s and a 2.2x myself, I'm trying to understand why you would choose the parametric equalizer function over the use of plotting your own target curves, whether there is an audible benefit doing so.
                                            >
                                            > FYI, we 2.2x model users have a way to bypass correction of any frequency range by independently tracing the left and right target curves to that of the measurement curves.
                                            >
                                            > Joseph
                                            >
                                            >
                                            > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                            > >
                                            > > My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000. So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                            > >
                                            > > Yes, near-field listening reduces room effects heard at the listening position as to high frequencies. But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s. While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well. In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up. Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                            > >
                                            > > FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters. It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency. Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.
                                            > >
                                            > > >>> "uohh" <uohh@> 10/7/2010 5:09 PM >>>
                                            > > By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.
                                            > >
                                            > > And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.
                                            > >
                                            > >
                                            > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                            > > >
                                            > > > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT. I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years.
                                            > > >
                                            > > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters. Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                            > > >
                                            > > > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies. Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening. I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics. With DSP, it's all just number crunching.
                                            > > >
                                            > > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                            > > > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                            > > >
                                            > > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                            > > > >
                                            > > > > From the current Zequalizer manual:
                                            > > > >
                                            > > > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                            > > > >
                                            > > > > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                            > > > >
                                            > > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                            > > > >
                                            > > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                            > > > > * His claim is that the thing that needs EQ (though NOT to flat)
                                            > > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                            > > > > This can only be obtained by using dummy head measurement
                                            > > > > or even better measurement on your own head.
                                            > > > > Alternatively--and this it the REAL POINT--you should do this
                                            > > > > by listening.*
                                            > > > >
                                            > > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                            > > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                            > > > > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                                            > > > > if you use just one speaker at a time?
                                            > > > >
                                            > > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                            > > > > small microphone that is placed just outside the ear canal on a little
                                            > > > > stalk.
                                            > > > >
                                            > > > > Also with something like the Z-Systems is there a typical Q for the filters
                                            > > > > doing the correction?
                                            > > > >
                                            > > > > Richard
                                            > > > >
                                            > > > > -----Original Message-----
                                            > > > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                            > > > > On Behalf Of Robert
                                            > > > > Sent: 06 October 2010 02:26
                                            > > > > To: regsaudioforum@yahoogroups.com
                                            > > > > Subject: [regsaudioforum] Listening and engineers
                                            > > > >
                                            > > > >
                                            > > > >
                                            > > > > I think we are somehow not getting the message of
                                            > > > > Staffelt's article, probably because I have not made
                                            > > > > it clear.
                                            > > > >
                                            > > > > According to his article, NO system that measures
                                            > > > > response in the abstract(in the absence of the head)
                                            > > > > is going to work right.
                                            > > > >
                                            > > > > His claim is that the thing that needs EQ(though NOT to flat)
                                            > > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                            > > > > This can only be obtained by using dummy head measurement
                                            > > > > or even better measurement on your own head.
                                            > > > > Alternatively--and this it the REAL POINT--you should do this
                                            > > > > by listening.
                                            > > > >
                                            > > > > Almost everyone is ultimately disappointed in almost all the
                                            > > > > automatic systems, except the ones which have detailed user
                                            > > > > control.
                                            > > > >
                                            > > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                            > > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                            > > > >
                                            > > > > Engineers do not like this, philoisophically. It seems unscientific
                                            > > > > and "subjective" to them. But to me it makes perfect sense and
                                            > > > > also fits with my experience.
                                            > > > >
                                            > > > > I have really discovered that I do not like any automatic system
                                            > > > > for the finishing touches. ALL the usual models of how we hear sound in
                                            > > > > rooms are wrong, except the ones that are so complicated that one cannot use
                                            > > > > them or the ones that measure with your head in place.
                                            > > > > They have to be wrong, actually because of how complicated things are. A
                                            > > > > correct model would be a lot more complicated than a single frequency
                                            > > > > varying time window for example.
                                            > > > >
                                            > > > > But none of those latter kind(dummy head or your own head) are offered for
                                            > > > > sale.
                                            > > > > Every system on the market measures the sound in the absence of the head.
                                            > > > > And almost all of them try to deal with things by using time windows.
                                            > > > > According to Staffelt this won't work and in my experience it does not work,
                                            > > > > not exactly.
                                            > > > >
                                            > > > > None of this automatic stuff really works very well above the bass. But this
                                            > > > > is not to say that SOMETHING does not work well above the bass. As I used to
                                            > > > > say to GF, since you can change the sound in the higher frequencies you are
                                            > > > > almost bound to be able to make it better!
                                            > > > >
                                            > > > > The problem is the same as audio has always had. Engineers like engineering
                                            > > > > solutions, and solutions that involve human experience as an evaluation tool
                                            > > > > make engineers nervous, in general.
                                            > > > >
                                            > > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                            > > > > right. Of course you have somehow to find out what right means, as TM
                                            > > > > suggested. But you can experiment and learn to hear what is right quite
                                            > > > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                            > > > > think ought to be good and set up something that makes them sound right in
                                            > > > > some intuitive sense.
                                            > > > >
                                            > > > > This could take some time! But it beats messing around trying to do the same
                                            > > > > thing by buying different equipment items or modifying your room. And it is
                                            > > > > a lot cheaper too.
                                            > > > >
                                            > > > > Of course one needs some sort of half decent listening room to begin with.
                                            > > > > But that is not so hard to arrange.
                                            > > > >
                                            > > > > I think the big failing here is human--people have not worked with user
                                            > > > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                            > > > > rather buy stuff than work even a little.
                                            > > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                            > > > > to play with EQ.
                                            > > > >
                                            > > > > I find it really intriguing to try to adjust one speaker to sound like
                                            > > > > another. Interesting to try!
                                            > > > >
                                            > > > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                            > > > > be. Just look at those NRC measurements. Forget questions of what radiation
                                            > > > > pattern ought to be for the moment and just look at one axis response.
                                            > > > > ALmost NO speaker is as smooth as it ought to be.
                                            > > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                            > > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                            > > > > ought to be(independently of what "flat " means, everyone agrees that things
                                            > > > > ought to be really smooth--but they are not, not in detail. Even getting
                                            > > > > totally smooth 1/3 octave is quite unusual without EQ).
                                            > > > >
                                            > > > > Anyway, anything with a name like Autocal is probably not really a total
                                            > > > > soultion above the bass --because if it is Auto it is going to produce
                                            > > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                            > > > > listening.(I did not watch long enough to find out whether Autocal works
                                            > > > > above the bass--the name is just an example).
                                            > > > >
                                            > > > > Engineers don't like this idea and most of them do not trust the customers
                                            > > > > either. But it is your music--and your head! whose transfer function you
                                            > > > > listen to.
                                            > > > >
                                            > > > > REG
                                            > > > >
                                            > > > >
                                            > > > >
                                            > > > >
                                            > > > > ------------------------------------
                                            > > > >
                                            > > > > Yahoo! Groups Links
                                            > > > >
                                            > > > >
                                            > > > >
                                            > > > >
                                            > > > >
                                            > > > > ------------------------------------
                                            > > > >
                                            > > > > Yahoo! Groups Links
                                            > > > >
                                            > > >
                                            > > >
                                            > > >
                                            > > >
                                            > > > ------------------------------------
                                            > > >
                                            > > > Yahoo! Groups Links
                                            > > >
                                            > >
                                            > >
                                            > >
                                            > >
                                            > > ------------------------------------
                                            > >
                                            > > Yahoo! Groups Links
                                            > >
                                            >
                                            >
                                            >
                                            >
                                            > ------------------------------------
                                            >
                                            > Yahoo! Groups Links
                                            >
                                          • Robert
                                            They do indeed use high resolution in the frequency domain. No one has ever responded to my graphs showing that the response is highly variable at this level
                                            Message 21 of 24 , Oct 9, 2010
                                            • 0 Attachment
                                              They do indeed use high resolution in the frequency domain.
                                              No one has ever responded to my graphs showing that the response is highly variable at this level of resolution with spatial position.

                                              This of course means that high resolution correction makes rather little sense.

                                              One intelligent approach is to correct the speakers themselves
                                              (after buying some that are plausibly behaved in pattern)
                                              and do room correction with farily low resolution in the base.

                                              This is why the Essex device was potentially so terrific.
                                              http://www.regonaudio.com/Arion%20Essex.html

                                              It got the speakers right and then all one had to do was fix
                                              the bass at something like the 1/6th octave EQ level and bingo.

                                              REG
                                              REG

                                              --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                              >
                                              > My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000. So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                              >
                                              > Yes, near-field listening reduces room effects heard at the listening position as to high frequencies. But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s. While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well. In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up. Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                              >
                                              > FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters. It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency. Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.
                                              >
                                              > >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                              > By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.
                                              >
                                              > And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.
                                              >
                                              >
                                              > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                              > >
                                              > > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT. I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years.
                                              > >
                                              > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters. Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                              > >
                                              > > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies. Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening. I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics. With DSP, it's all just number crunching.
                                              > >
                                              > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                              > > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                              > >
                                              > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                              > > >
                                              > > > From the current Zequalizer manual:
                                              > > >
                                              > > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                              > > >
                                              > > > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                              > > >
                                              > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                              > > >
                                              > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                              > > > * His claim is that the thing that needs EQ (though NOT to flat)
                                              > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                              > > > This can only be obtained by using dummy head measurement
                                              > > > or even better measurement on your own head.
                                              > > > Alternatively--and this it the REAL POINT--you should do this
                                              > > > by listening.*
                                              > > >
                                              > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                              > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                              > > > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                                              > > > if you use just one speaker at a time?
                                              > > >
                                              > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                              > > > small microphone that is placed just outside the ear canal on a little
                                              > > > stalk.
                                              > > >
                                              > > > Also with something like the Z-Systems is there a typical Q for the filters
                                              > > > doing the correction?
                                              > > >
                                              > > > Richard
                                              > > >
                                              > > > -----Original Message-----
                                              > > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                              > > > On Behalf Of Robert
                                              > > > Sent: 06 October 2010 02:26
                                              > > > To: regsaudioforum@yahoogroups.com
                                              > > > Subject: [regsaudioforum] Listening and engineers
                                              > > >
                                              > > >
                                              > > >
                                              > > > I think we are somehow not getting the message of
                                              > > > Staffelt's article, probably because I have not made
                                              > > > it clear.
                                              > > >
                                              > > > According to his article, NO system that measures
                                              > > > response in the abstract(in the absence of the head)
                                              > > > is going to work right.
                                              > > >
                                              > > > His claim is that the thing that needs EQ(though NOT to flat)
                                              > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                              > > > This can only be obtained by using dummy head measurement
                                              > > > or even better measurement on your own head.
                                              > > > Alternatively--and this it the REAL POINT--you should do this
                                              > > > by listening.
                                              > > >
                                              > > > Almost everyone is ultimately disappointed in almost all the
                                              > > > automatic systems, except the ones which have detailed user
                                              > > > control.
                                              > > >
                                              > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                              > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                              > > >
                                              > > > Engineers do not like this, philoisophically. It seems unscientific
                                              > > > and "subjective" to them. But to me it makes perfect sense and
                                              > > > also fits with my experience.
                                              > > >
                                              > > > I have really discovered that I do not like any automatic system
                                              > > > for the finishing touches. ALL the usual models of how we hear sound in
                                              > > > rooms are wrong, except the ones that are so complicated that one cannot use
                                              > > > them or the ones that measure with your head in place.
                                              > > > They have to be wrong, actually because of how complicated things are. A
                                              > > > correct model would be a lot more complicated than a single frequency
                                              > > > varying time window for example.
                                              > > >
                                              > > > But none of those latter kind(dummy head or your own head) are offered for
                                              > > > sale.
                                              > > > Every system on the market measures the sound in the absence of the head.
                                              > > > And almost all of them try to deal with things by using time windows.
                                              > > > According to Staffelt this won't work and in my experience it does not work,
                                              > > > not exactly.
                                              > > >
                                              > > > None of this automatic stuff really works very well above the bass. But this
                                              > > > is not to say that SOMETHING does not work well above the bass. As I used to
                                              > > > say to GF, since you can change the sound in the higher frequencies you are
                                              > > > almost bound to be able to make it better!
                                              > > >
                                              > > > The problem is the same as audio has always had. Engineers like engineering
                                              > > > solutions, and solutions that involve human experience as an evaluation tool
                                              > > > make engineers nervous, in general.
                                              > > >
                                              > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                              > > > right. Of course you have somehow to find out what right means, as TM
                                              > > > suggested. But you can experiment and learn to hear what is right quite
                                              > > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                              > > > think ought to be good and set up something that makes them sound right in
                                              > > > some intuitive sense.
                                              > > >
                                              > > > This could take some time! But it beats messing around trying to do the same
                                              > > > thing by buying different equipment items or modifying your room. And it is
                                              > > > a lot cheaper too.
                                              > > >
                                              > > > Of course one needs some sort of half decent listening room to begin with.
                                              > > > But that is not so hard to arrange.
                                              > > >
                                              > > > I think the big failing here is human--people have not worked with user
                                              > > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                              > > > rather buy stuff than work even a little.
                                              > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                              > > > to play with EQ.
                                              > > >
                                              > > > I find it really intriguing to try to adjust one speaker to sound like
                                              > > > another. Interesting to try!
                                              > > >
                                              > > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                              > > > be. Just look at those NRC measurements. Forget questions of what radiation
                                              > > > pattern ought to be for the moment and just look at one axis response.
                                              > > > ALmost NO speaker is as smooth as it ought to be.
                                              > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                              > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                              > > > ought to be(independently of what "flat " means, everyone agrees that things
                                              > > > ought to be really smooth--but they are not, not in detail. Even getting
                                              > > > totally smooth 1/3 octave is quite unusual without EQ).
                                              > > >
                                              > > > Anyway, anything with a name like Autocal is probably not really a total
                                              > > > soultion above the bass --because if it is Auto it is going to produce
                                              > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                              > > > listening.(I did not watch long enough to find out whether Autocal works
                                              > > > above the bass--the name is just an example).
                                              > > >
                                              > > > Engineers don't like this idea and most of them do not trust the customers
                                              > > > either. But it is your music--and your head! whose transfer function you
                                              > > > listen to.
                                              > > >
                                              > > > REG
                                              > > >
                                              > > >
                                              > > >
                                              > > >
                                              > > > ------------------------------------
                                              > > >
                                              > > > Yahoo! Groups Links
                                              > > >
                                              > > >
                                              > > >
                                              > > >
                                              > > >
                                              > > > ------------------------------------
                                              > > >
                                              > > > Yahoo! Groups Links
                                              > > >
                                              > >
                                              > >
                                              > >
                                              > >
                                              > > ------------------------------------
                                              > >
                                              > > Yahoo! Groups Links
                                              > >
                                              >
                                              >
                                              >
                                              >
                                              > ------------------------------------
                                              >
                                              > Yahoo! Groups Links
                                              >
                                            • ymm
                                              Someone should alert AS to his manufacturing of a bass correctable M40.2, like the B and O flagship speaker. That would be a really great speaker for almost
                                              Message 22 of 24 , Oct 10, 2010
                                              • 0 Attachment
                                                Someone should alert AS to his manufacturing of  a bass correctable M40.2, like the B and O flagship speaker. 

                                                That would be a really great speaker for almost all reasonable rooms

                                                Yip 


                                                From: Robert <regonaudio@...>
                                                To: regsaudioforum@yahoogroups.com
                                                Sent: Sunday, 10 October 2010 11:27:32
                                                Subject: [regsaudioforum] Re: Listening and engineers

                                                 

                                                They do indeed use high resolution in the frequency domain.
                                                No one has ever responded to my graphs showing that the response is highly variable at this level of resolution with spatial position.

                                                This of course means that high resolution correction makes rather little sense.

                                                One intelligent approach is to correct the speakers themselves
                                                (after buying some that are plausibly behaved in pattern)
                                                and do room correction with farily low resolution in the base.

                                                This is why the Essex device was potentially so terrific.
                                                http://www.regonaudio.com/Arion%20Essex.html

                                                It got the speakers right and then all one had to do was fix
                                                the bass at something like the 1/6th octave EQ level and bingo.

                                                REG
                                                REG

                                                --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                                >
                                                > My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000. So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                                >
                                                > Yes, near-field listening reduces room effects heard at the listening position as to high frequencies. But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s. While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well. In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up. Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                                >
                                                > FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters. It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency. Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.
                                                >
                                                > >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                                > By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.
                                                >
                                                > And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.
                                                >
                                                >
                                                > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                > >
                                                > > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT. I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years.
                                                > >
                                                > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters. Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                                > >
                                                > > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies. Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening. I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics. With DSP, it's all just number crunching.
                                                > >
                                                > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                                > > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                                > >
                                                > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                > > >
                                                > > > From the current Zequalizer manual:
                                                > > >
                                                > > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                                > > >
                                                > > > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                                > > >
                                                > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                                > > >
                                                > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                                > > > * His claim is that the thing that needs EQ (though NOT to flat)
                                                > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                                > > > This can only be obtained by using dummy head measurement
                                                > > > or even better measurement on your own head.
                                                > > > Alternatively--and this it the REAL POINT--you should do this
                                                > > > by listening.*
                                                > > >
                                                > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                                > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                                > > > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                                                > > > if you use just one speaker at a time?
                                                > > >
                                                > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                                > > > small microphone that is placed just outside the ear canal on a little
                                                > > > stalk.
                                                > > >
                                                > > > Also with something like the Z-Systems is there a typical Q for the filters
                                                > > > doing the correction?
                                                > > >
                                                > > > Richard
                                                > > >
                                                > > > -----Original Message-----
                                                > > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                                > > > On Behalf Of Robert
                                                > > > Sent: 06 October 2010 02:26
                                                > > > To: regsaudioforum@yahoogroups.com
                                                > > > Subject: [regsaudioforum] Listening and engineers
                                                > > >
                                                > > >
                                                > > >
                                                > > > I think we are somehow not getting the message of
                                                > > > Staffelt's article, probably because I have not made
                                                > > > it clear.
                                                > > >
                                                > > > According to his article, NO system that measures
                                                > > > response in the abstract(in the absence of the head)
                                                > > > is going to work right.
                                                > > >
                                                > > > His claim is that the thing that needs EQ(though NOT to flat)
                                                > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                                > > > This can only be obtained by using dummy head measurement
                                                > > > or even better measurement on your own head.
                                                > > > Alternatively--and this it the REAL POINT--you should do this
                                                > > > by listening.
                                                > > >
                                                > > > Almost everyone is ultimately disappointed in almost all the
                                                > > > automatic systems, except the ones which have detailed user
                                                > > > control.
                                                > > >
                                                > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                                > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                                > > >
                                                > > > Engineers do not like this, philoisophically. It seems unscientific
                                                > > > and "subjective" to them. But to me it makes perfect sense and
                                                > > > also fits with my experience.
                                                > > >
                                                > > > I have really discovered that I do not like any automatic system
                                                > > > for the finishing touches. ALL the usual models of how we hear sound in
                                                > > > rooms are wrong, except the ones that are so complicated that one cannot use
                                                > > > them or the ones that measure with your head in place.
                                                > > > They have to be wrong, actually because of how complicated things are. A
                                                > > > correct model would be a lot more complicated than a single frequency
                                                > > > varying time window for example.
                                                > > >
                                                > > > But none of those latter kind(dummy head or your own head) are offered for
                                                > > > sale.
                                                > > > Every system on the market measures the sound in the absence of the head.
                                                > > > And almost all of them try to deal with things by using time windows.
                                                > > > According to Staffelt this won't work and in my experience it does not work,
                                                > > > not exactly.
                                                > > >
                                                > > > None of this automatic stuff really works very well above the bass. But this
                                                > > > is not to say that SOMETHING does not work well above the bass. As I used to
                                                > > > say to GF, since you can change the sound in the higher frequencies you are
                                                > > > almost bound to be able to make it better!
                                                > > >
                                                > > > The problem is the same as audio has always had. Engineers like engineering
                                                > > > solutions, and solutions that involve human experience as an evaluation tool
                                                > > > make engineers nervous, in general.
                                                > > >
                                                > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                                > > > right. Of course you have somehow to find out what right means, as TM
                                                > > > suggested. But you can experiment and learn to hear what is right quite
                                                > > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                                > > > think ought to be good and set up something that makes them sound right in
                                                > > > some intuitive sense.
                                                > > >
                                                > > > This could take some time! But it beats messing around trying to do the same
                                                > > > thing by buying different equipment items or modifying your room. And it is
                                                > > > a lot cheaper too.
                                                > > >
                                                > > > Of course one needs some sort of half decent listening room to begin with.
                                                > > > But that is not so hard to arrange.
                                                > > >
                                                > > > I think the big failing here is human--people have not worked with user
                                                > > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                                > > > rather buy stuff than work even a little.
                                                > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                                > > > to play with EQ.
                                                > > >
                                                > > > I find it really intriguing to try to adjust one speaker to sound like
                                                > > > another. Interesting to try!
                                                > > >
                                                > > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                                > > > be. Just look at those NRC measurements. Forget questions of what radiation
                                                > > > pattern ought to be for the moment and just look at one axis response.
                                                > > > ALmost NO speaker is as smooth as it ought to be.
                                                > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                                > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                                > > > ought to be(independently of what "flat " means, everyone agrees that things
                                                > > > ought to be really smooth--but they are not, not in detail. Even getting
                                                > > > totally smooth 1/3 octave is quite unusual without EQ).
                                                > > >
                                                > > > Anyway, anything with a name like Autocal is probably not really a total
                                                > > > soultion above the bass --because if it is Auto it is going to produce
                                                > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                                > > > listening.(I did not watch long enough to find out whether Autocal works
                                                > > > above the bass--the name is just an example).
                                                > > >
                                                > > > Engineers don't like this idea and most of them do not trust the customers
                                                > > > either. But it is your music--and your head! whose transfer function you
                                                > > > listen to.
                                                > > >
                                                > > > REG
                                                > > >
                                                > > >
                                                > > >
                                                > > >
                                                > > > ------------------------------------
                                                > > >
                                                > > > Yahoo! Groups Links
                                                > > >
                                                > > >
                                                > > >
                                                > > >
                                                > > >
                                                > > > ------------------------------------
                                                > > >
                                                > > > Yahoo! Groups Links
                                                > > >
                                                > >
                                                > >
                                                > >
                                                > >
                                                > > ------------------------------------
                                                > >
                                                > > Yahoo! Groups Links
                                                > >
                                                >
                                                >
                                                >
                                                >
                                                > ------------------------------------
                                                >
                                                > Yahoo! Groups Links
                                                >


                                              • Robert
                                                This is easily done as an addon. There is really no reason at all to put this IN THE SPEAKER. REG
                                                Message 23 of 24 , Oct 10, 2010
                                                • 0 Attachment
                                                  This is easily done as an addon.
                                                  There is really no reason at all to put this
                                                  IN THE SPEAKER.

                                                  REG

                                                  --- In regsaudioforum@yahoogroups.com, ymm <yipmangmeng@...> wrote:
                                                  >
                                                  > Someone should alert AS to his manufacturing of a bass correctable M40.2, like
                                                  > the B and O flagship speaker.
                                                  >
                                                  > That would be a really great speaker for almost all reasonable rooms
                                                  >
                                                  > Yip
                                                  >
                                                  >
                                                  >
                                                  > ________________________________
                                                  > From: Robert <regonaudio@...>
                                                  > To: regsaudioforum@yahoogroups.com
                                                  > Sent: Sunday, 10 October 2010 11:27:32
                                                  > Subject: [regsaudioforum] Re: Listening and engineers
                                                  >
                                                  >
                                                  > They do indeed use high resolution in the frequency domain.
                                                  > No one has ever responded to my graphs showing that the response is highly
                                                  > variable at this level of resolution with spatial position.
                                                  >
                                                  > This of course means that high resolution correction makes rather little sense.
                                                  >
                                                  > One intelligent approach is to correct the speakers themselves
                                                  > (after buying some that are plausibly behaved in pattern)
                                                  > and do room correction with farily low resolution in the base.
                                                  >
                                                  > This is why the Essex device was potentially so terrific.
                                                  > http://www.regonaudio.com/Arion%20Essex.html
                                                  >
                                                  > It got the speakers right and then all one had to do was fix
                                                  > the bass at something like the 1/6th octave EQ level and bingo.
                                                  >
                                                  > REG
                                                  > REG
                                                  >
                                                  > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                  > >
                                                  > > My understanding is that, apart from Acourate, devices like the TacT and the
                                                  > >Lyngdorf correct the response using the equivalent of more individual filters
                                                  > >than any other, more than 1,000. So while there may be smoothing (1/10-octave,
                                                  > >I've heard) going on, there is a less of it than with 12 parametric filters or
                                                  > >31 1/3-octave filters.
                                                  > >
                                                  > > Yes, near-field listening reduces room effects heard at the listening position
                                                  > >as to high frequencies. But with good physical positioning and good room
                                                  > >treatment this makes high-frequency response correction less necessary, assuming
                                                  > >you like what you hear from the speakers, which is certainly the case with my
                                                  > >M40.1s. While I have heard the TacT sharpen the imaging and stage of my system
                                                  > >in other set ups, in my current set up the combination of fairly exact physical
                                                  > >positioning, good room treatment, near field listening, and good speaker
                                                  > >matching by Harbeth means that from about 1 kHz on up the TacT measurements for
                                                  > >the left and right channel overlay each other quite well. In addition, my most
                                                  > >recent TacT measurement showed a computed delay for the left speaker of 8.73
                                                  > >milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty
                                                  > >symmetrical set up. Perhaps for these reasons I don't hear any image/staging
                                                  > >enhancement when engaging the TacT target curve function full range compared to
                                                  > >listening through the TacT in its bypass mode.
                                                  > >
                                                  > > FYI, the time delay the TacT can insert to compensate for unequal air path
                                                  > >lengths from the speakers to the listening position can be used or not used
                                                  > >independent of the target curve correction or the parametric filters. It is
                                                  > >also manually adjustable to compensate for factors other than air path length
                                                  > >differences, such as digital filter latency. Thus, for example, you don't have
                                                  > >to move your subwoofers a meter out of the corners to compensate for the latency
                                                  > >of the filters, as with the subwoofer response correcting device which REG
                                                  > >recently reviewed in TAS.
                                                  > >
                                                  > > >>> "uohh" <uohh@> 10/7/2010 5:09 PM >>>
                                                  > > By using target curve, does it really make it more micro-correcting than other
                                                  > >types of equalizers? My understanding is that Tact uses a lot of smoothing on
                                                  > >it's measurement curves, so if you apply a target curve to it, it really is, in
                                                  > >effect, a broad-brush change.
                                                  > >
                                                  > > And I think the other major benefit of using a target curve, on both channels'
                                                  > >higher frequency region, is the matching of the FR of the two speakers. With
                                                  > >nearfield listening position and short window measurement, the resulting
                                                  > >correction should be more of a speaker adjustment than a room sound correction.
                                                  > >
                                                  > >
                                                  > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                  > > >
                                                  > > > I've owned all different types of equalizers, from the Dick-Burwen-inspired
                                                  > >Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and
                                                  > >Audient graphics, and currently the TacT. I also used professional sound desk
                                                  > >parametric and graphic EQ for years as part of volunteer live sound engineering
                                                  > >for a large church for about ten years.
                                                  > >
                                                  > > >
                                                  > > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a
                                                  > >micro-correction target curve approach and a more broad-brush approach like the
                                                  > >use of parametric filters. Micro-correction might be best in the bass region
                                                  > >since in that region what you see is more or less what you sonically get.
                                                  > >
                                                  > > >
                                                  > > > REG's discussion of the Staffelt research seems to suggest that a broad brush
                                                  > >approach might be subjectively better, at least in the higher frequencies. Yes,
                                                  > >it's expensive, but a unit like the TacT give you choices and those choices can
                                                  > >be implemented on the fly while you are listening. I've asked TacT to consider
                                                  > >swapping the parametrics for 1/3-octave graphics, or even adding the graphic
                                                  > >capability in addition to the parametrics. With DSP, it's all just number
                                                  > >crunching.
                                                  > > >
                                                  > > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                                  > > > With the ability to choose and shape a target curve, why do you want to use
                                                  > >the equalizer and mess with Q and unwanted changes in other regions?
                                                  > > >
                                                  > > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                  > > > >
                                                  > > > > From the current Zequalizer manual:
                                                  > > > >
                                                  > > > > "Q is the inverse of bandwidth. It is the product of center frequency
                                                  > >divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide
                                                  > >curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth
                                                  > >of a typical midrange EQ in an analog equalizer."
                                                  > > > >
                                                  > > > > The Z-Systems EQ products have parametric equalizers with widely variable
                                                  > >Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to
                                                  > >a 1/3-octave graphic equalizer.
                                                  > > > >
                                                  > > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction
                                                  > >of an octave you wish it to cover in .1 octave increments. Thus, you could set
                                                  > >it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave
                                                  > >correction. With the latest V 1.0 software, you can use the target curve
                                                  > >function to correct the bass up to a frequency of your choosing, say 700 to 1000
                                                  > >Hz with no target curve correction above the frequency you choose. Above that
                                                  > >frequency, you could use the parametrics and have just about enough of them to
                                                  > >cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                                  > > > >
                                                  > > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                                  > > > > * His claim is that the thing that needs EQ (though NOT to flat)
                                                  > > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                                  > > > > This can only be obtained by using dummy head measurement
                                                  > > > > or even better measurement on your own head.
                                                  > > > > Alternatively--and this it the REAL POINT--you should do this
                                                  > > > > by listening.*
                                                  > > > >
                                                  > > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                                  > > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like
                                                  > >Sensaura
                                                  > > > > that uses the HRTF to deal with inter-channel crosstalk. Does it get
                                                  > >easier
                                                  > > > > if you use just one speaker at a time?
                                                  > > > >
                                                  > > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                                  > > > > small microphone that is placed just outside the ear canal on a little
                                                  > > > > stalk.
                                                  > > > >
                                                  > > > > Also with something like the Z-Systems is there a typical Q for the
                                                  > filters
                                                  > > > > doing the correction?
                                                  > > > >
                                                  > > > > Richard
                                                  > > > >
                                                  > > > > -----Original Message-----
                                                  > > > > From: regsaudioforum@yahoogroups.com
                                                  > >[mailto:regsaudioforum@yahoogroups.com]
                                                  > >
                                                  > > > > On Behalf Of Robert
                                                  > > > > Sent: 06 October 2010 02:26
                                                  > > > > To: regsaudioforum@yahoogroups.com
                                                  > > > > Subject: [regsaudioforum] Listening and engineers
                                                  > > > >
                                                  > > > >
                                                  > > > >
                                                  > > > > I think we are somehow not getting the message of
                                                  > > > > Staffelt's article, probably because I have not made
                                                  > > > > it clear.
                                                  > > > >
                                                  > > > > According to his article, NO system that measures
                                                  > > > > response in the abstract(in the absence of the head)
                                                  > > > > is going to work right.
                                                  > > > >
                                                  > > > > His claim is that the thing that needs EQ(though NOT to flat)
                                                  > > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                                  > > > > This can only be obtained by using dummy head measurement
                                                  > > > > or even better measurement on your own head.
                                                  > > > > Alternatively--and this it the REAL POINT--you should do this
                                                  > > > > by listening.
                                                  > > > >
                                                  > > > > Almost everyone is ultimately disappointed in almost all the
                                                  > > > > automatic systems, except the ones which have detailed user
                                                  > > > > control.
                                                  > > > >
                                                  > > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                                  > > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                                  > > > >
                                                  > > > > Engineers do not like this, philoisophically. It seems unscientific
                                                  > > > > and "subjective" to them. But to me it makes perfect sense and
                                                  > > > > also fits with my experience.
                                                  > > > >
                                                  > > > > I have really discovered that I do not like any automatic system
                                                  > > > > for the finishing touches. ALL the usual models of how we hear sound in
                                                  > > > > rooms are wrong, except the ones that are so complicated that one cannot
                                                  > >use
                                                  > > > > them or the ones that measure with your head in place.
                                                  > > > > They have to be wrong, actually because of how complicated things are. A
                                                  > > > > correct model would be a lot more complicated than a single frequency
                                                  > > > > varying time window for example.
                                                  > > > >
                                                  > > > > But none of those latter kind(dummy head or your own head) are offered for
                                                  > > > > sale.
                                                  > > > > Every system on the market measures the sound in the absence of the head.
                                                  > > > > And almost all of them try to deal with things by using time windows.
                                                  > > > > According to Staffelt this won't work and in my experience it does not
                                                  > >work,
                                                  > > > > not exactly.
                                                  > > > >
                                                  > > > > None of this automatic stuff really works very well above the bass. But
                                                  > >this
                                                  > > > > is not to say that SOMETHING does not work well above the bass. As I used
                                                  > >to
                                                  > > > > say to GF, since you can change the sound in the higher frequencies you
                                                  > are
                                                  > > > > almost bound to be able to make it better!
                                                  > > > >
                                                  > > > > The problem is the same as audio has always had. Engineers like
                                                  > engineering
                                                  > > > > solutions, and solutions that involve human experience as an evaluation
                                                  > >tool
                                                  > > > > make engineers nervous, in general.
                                                  > > > >
                                                  > > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                                  > > > > right. Of course you have somehow to find out what right means, as TM
                                                  > > > > suggested. But you can experiment and learn to hear what is right quite
                                                  > > > > easily. Just try a lot of EQs both on pink noise and on recordings that
                                                  > you
                                                  > > > > think ought to be good and set up something that makes them sound right in
                                                  > > > > some intuitive sense.
                                                  > > > >
                                                  > > > > This could take some time! But it beats messing around trying to do the
                                                  > >same
                                                  > > > > thing by buying different equipment items or modifying your room. And it
                                                  > >is
                                                  > > > > a lot cheaper too.
                                                  > > > >
                                                  > > > > Of course one needs some sort of half decent listening room to begin with.
                                                  > > > > But that is not so hard to arrange.
                                                  > > > >
                                                  > > > > I think the big failing here is human--people have not worked with user
                                                  > > > > controlled EQ carefully enough. Talk about laziness! People would
                                                  > >apparently
                                                  > > > > rather buy stuff than work even a little.
                                                  > > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                                  > > > > to play with EQ.
                                                  > > > >
                                                  > > > > I find it really intriguing to try to adjust one speaker to sound like
                                                  > > > > another. Interesting to try!
                                                  > > > >
                                                  > > > > And don't forget that NO speaker, essentially , is as smooth as it ought
                                                  > to
                                                  > > > > be. Just look at those NRC measurements. Forget questions of what
                                                  > radiation
                                                  > > > > pattern ought to be for the moment and just look at one axis response.
                                                  > > > > ALmost NO speaker is as smooth as it ought to be.
                                                  > > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                                  > > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                                  > > > > ought to be(independently of what "flat " means, everyone agrees that
                                                  > >things
                                                  > > > > ought to be really smooth--but they are not, not in detail. Even getting
                                                  > > > > totally smooth 1/3 octave is quite unusual without EQ).
                                                  > > > >
                                                  > > > > Anyway, anything with a name like Autocal is probably not really a total
                                                  > > > > soultion above the bass --because if it is Auto it is going to produce
                                                  > > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                                  > > > > listening.(I did not watch long enough to find out whether Autocal works
                                                  > > > > above the bass--the name is just an example).
                                                  > > > >
                                                  > > > > Engineers don't like this idea and most of them do not trust the customers
                                                  > > > > either. But it is your music--and your head! whose transfer function you
                                                  > > > > listen to.
                                                  > > > >
                                                  > > > > REG
                                                  > > > >
                                                  > > > >
                                                  > > > >
                                                  > > > >
                                                  > > > > ------------------------------------
                                                  > > > >
                                                  > > > > Yahoo! Groups Links
                                                  > > > >
                                                  > > > >
                                                  > > > >
                                                  > > > >
                                                  > > > >
                                                  > > > > ------------------------------------
                                                  > > > >
                                                  > > > > Yahoo! Groups Links
                                                  > > > >
                                                  > > >
                                                  > > >
                                                  > > >
                                                  > > >
                                                  > > > ------------------------------------
                                                  > > >
                                                  > > > Yahoo! Groups Links
                                                  > > >
                                                  > >
                                                  > >
                                                  > >
                                                  > >
                                                  > > ------------------------------------
                                                  > >
                                                  > > Yahoo! Groups Links
                                                  > >
                                                  >
                                                • Will_H
                                                  I suggested that to him when the M40.1 was in development... ... From: ymm To: regsaudioforum@yahoogroups.com Sent: Sunday, October 10, 2010 5:39 AM Subject:
                                                  Message 24 of 24 , Oct 10, 2010
                                                  • 0 Attachment
                                                    
                                                    I suggested that to him when the M40.1 was in development...
                                                     
                                                     
                                                     
                                                    ----- Original Message -----
                                                    From: ymm
                                                    Sent: Sunday, October 10, 2010 5:39 AM
                                                    Subject: Re: [regsaudioforum] Re: Listening and engineers

                                                    Someone should alert AS to his manufacturing of  a bass correctable M40.2, like the B and O flagship speaker. 

                                                    That would be a really great speaker for almost all reasonable rooms

                                                    Yip 


                                                    From: Robert <regonaudio@...>
                                                    To: regsaudioforum@yahoogroups.com
                                                    Sent: Sunday, 10 October 2010 11:27:32
                                                    Subject: [regsaudioforum] Re: Listening and engineers

                                                     

                                                    They do indeed use high resolution in the frequency domain.
                                                    No one has ever responded to my graphs showing that the response is highly variable at this level of resolution with spatial position.

                                                    This of course means that high resolution correction makes rather little sense.

                                                    One intelligent approach is to correct the speakers themselves
                                                    (after buying some that are plausibly behaved in pattern)
                                                    and do room correction with farily low resolution in the base.

                                                    This is why the Essex device was potentially so terrific.
                                                    http://www.regonaudio.com/Arion%20Essex.html

                                                    It got the speakers right and then all one had to do was fix
                                                    the bass at something like the 1/6th octave EQ level and bingo.

                                                    REG
                                                    REG

                                                    --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@...> wrote:
                                                    >
                                                    > My understanding is that, apart from Acourate, devices like the TacT and the Lyngdorf correct the response using the equivalent of more individual filters than any other, more than 1,000. So while there may be smoothing (1/10-octave, I've heard) going on, there is a less of it than with 12 parametric filters or 31 1/3-octave filters.
                                                    >
                                                    > Yes, near-field listening reduces room effects heard at the listening position as to high frequencies. But with good physical positioning and good room treatment this makes high-frequency response correction less necessary, assuming you like what you hear from the speakers, which is certainly the case with my M40.1s. While I have heard the TacT sharpen the imaging and stage of my system in other set ups, in my current set up the combination of fairly exact physical positioning, good room treatment, near field listening, and good speaker matching by Harbeth means that from about 1 kHz on up the TacT measurements for the left and right channel overlay each other quite well. In addition, my most recent TacT measurement showed a computed delay for the left speaker of 8.73 milliseconds and 8.74 milliseconds for the right speaker, indicating a pretty symmetrical set up. Perhaps for these reasons I don't hear any image/staging enhancement when engaging the TacT target curve function full range compared to listening through the TacT in its bypass mode.
                                                    >
                                                    > FYI, the time delay the TacT can insert to compensate for unequal air path lengths from the speakers to the listening position can be used or not used independent of the target curve correction or the parametric filters. It is also manually adjustable to compensate for factors other than air path length differences, such as digital filter latency. Thus, for example, you don't have to move your subwoofers a meter out of the corners to compensate for the latency of the filters, as with the subwoofer response correcting device which REG recently reviewed in TAS.
                                                    >
                                                    > >>> "uohh" <uohh@...> 10/7/2010 5:09 PM >>>
                                                    > By using target curve, does it really make it more micro-correcting than other types of equalizers? My understanding is that Tact uses a lot of smoothing on it's measurement curves, so if you apply a target curve to it, it really is, in effect, a broad-brush change.
                                                    >
                                                    > And I think the other major benefit of using a target curve, on both channels' higher frequency region, is the matching of the FR of the two speakers. With nearfield listening position and short window measurement, the resulting correction should be more of a speaker adjustment than a room sound correction.
                                                    >
                                                    >
                                                    > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                    > >
                                                    > > I've owned all different types of equalizers, from the Dick-Burwen-inspired Cello Palette Preamp, to the Z-Systems rdp-1, to the Rives PARC, to the Rane and Audient graphics, and currently the TacT. I also used professional sound desk parametric and graphic EQ for years as part of volunteer live sound engineering for a large church for about ten years.
                                                    > >
                                                    > > I'm just saying that the TacT RCS 2.2XP unit allows both use of a micro-correction target curve approach and a more broad-brush approach like the use of parametric filters. Micro-correction might be best in the bass region since in that region what you see is more or less what you sonically get.
                                                    > >
                                                    > > REG's discussion of the Staffelt research seems to suggest that a broad brush approach might be subjectively better, at least in the higher frequencies. Yes, it's expensive, but a unit like the TacT give you choices and those choices can be implemented on the fly while you are listening. I've asked TacT to consider swapping the parametrics for 1/3-octave graphics, or even adding the graphic capability in addition to the parametrics. With DSP, it's all just number crunching.
                                                    > >
                                                    > > >>> "uohh" <uohh@> 10/7/2010 3:24 PM >>>
                                                    > > With the ability to choose and shape a target curve, why do you want to use the equalizer and mess with Q and unwanted changes in other regions?
                                                    > >
                                                    > > --- In regsaudioforum@yahoogroups.com, "Tom Mallin" <tmallin@> wrote:
                                                    > > >
                                                    > > > From the current Zequalizer manual:
                                                    > > >
                                                    > > > "Q is the inverse of bandwidth. It is the product of center frequency divided by the 3 dB down bandwidth. Thus, a Q of 0.4 produces an extremely wide curve and will be rarely used. A Q of 0.6 or 0.7 corresponds with the bandwidth of a typical midrange EQ in an analog equalizer."
                                                    > > >
                                                    > > > The Z-Systems EQ products have parametric equalizers with widely variable Q. By my calculations, setting a Q value of 4.5 would be roughly equivalent to a 1/3-octave graphic equalizer.
                                                    > > >
                                                    > > > On the TacT, the 12-band parametric EQ can be set directly for the fraction of an octave you wish it to cover in .1 octave increments. Thus, you could set it to .3 or .4 octave bandwidth for a very rough equivalent of a 1/3-octave correction. With the latest V 1.0 software, you can use the target curve function to correct the bass up to a frequency of your choosing, say 700 to 1000 Hz with no target curve correction above the frequency you choose. Above that frequency, you could use the parametrics and have just about enough of them to cover the rest of the bandwidth with 1/3-octave wide filters adjusted by ear.
                                                    > > >
                                                    > > > >>> "Richard Tuck" <rtuck@> 10/7/2010 8:19 AM >>>
                                                    > > > * His claim is that the thing that needs EQ (though NOT to flat)
                                                    > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                                    > > > This can only be obtained by using dummy head measurement
                                                    > > > or even better measurement on your own head.
                                                    > > > Alternatively--and this it the REAL POINT--you should do this
                                                    > > > by listening.*
                                                    > > >
                                                    > > > I see a dilemma here, if I have got the right idea, "response AT THE
                                                    > > > ENTRANCE TO THE EAR CANAL" means using headphones or something like Sensaura
                                                    > > > that uses the HRTF to deal with inter-channel crosstalk. Does it get easier
                                                    > > > if you use just one speaker at a time?
                                                    > > >
                                                    > > > Maybe one can use one of these recently developed hearing aids with a VERY
                                                    > > > small microphone that is placed just outside the ear canal on a little
                                                    > > > stalk.
                                                    > > >
                                                    > > > Also with something like the Z-Systems is there a typical Q for the filters
                                                    > > > doing the correction?
                                                    > > >
                                                    > > > Richard
                                                    > > >
                                                    > > > -----Original Message-----
                                                    > > > From: regsaudioforum@yahoogroups.com [mailto:regsaudioforum@yahoogroups.com]
                                                    > > > On Behalf Of Robert
                                                    > > > Sent: 06 October 2010 02:26
                                                    > > > To: regsaudioforum@yahoogroups.com
                                                    > > > Subject: [regsaudioforum] Listening and engineers
                                                    > > >
                                                    > > >
                                                    > > >
                                                    > > > I think we are somehow not getting the message of
                                                    > > > Staffelt's article, probably because I have not made
                                                    > > > it clear.
                                                    > > >
                                                    > > > According to his article, NO system that measures
                                                    > > > response in the abstract(in the absence of the head)
                                                    > > > is going to work right.
                                                    > > >
                                                    > > > His claim is that the thing that needs EQ(though NOT to flat)
                                                    > > > is the (steady state)response AT THE ENTRANCE TO THE EAR CANAL.
                                                    > > > This can only be obtained by using dummy head measurement
                                                    > > > or even better measurement on your own head.
                                                    > > > Alternatively--and this it the REAL POINT--you should do this
                                                    > > > by listening.
                                                    > > >
                                                    > > > Almost everyone is ultimately disappointed in almost all the
                                                    > > > automatic systems, except the ones which have detailed user
                                                    > > > control.
                                                    > > >
                                                    > > > YOU REALLY NEED TO DO EVERYTHING EXCEPT BASS BY EAR.
                                                    > > > Automatic roughs it out, but after that, you need to fine tune by ear.
                                                    > > >
                                                    > > > Engineers do not like this, philoisophically. It seems unscientific
                                                    > > > and "subjective" to them. But to me it makes perfect sense and
                                                    > > > also fits with my experience.
                                                    > > >
                                                    > > > I have really discovered that I do not like any automatic system
                                                    > > > for the finishing touches. ALL the usual models of how we hear sound in
                                                    > > > rooms are wrong, except the ones that are so complicated that one cannot use
                                                    > > > them or the ones that measure with your head in place.
                                                    > > > They have to be wrong, actually because of how complicated things are. A
                                                    > > > correct model would be a lot more complicated than a single frequency
                                                    > > > varying time window for example.
                                                    > > >
                                                    > > > But none of those latter kind(dummy head or your own head) are offered for
                                                    > > > sale.
                                                    > > > Every system on the market measures the sound in the absence of the head.
                                                    > > > And almost all of them try to deal with things by using time windows.
                                                    > > > According to Staffelt this won't work and in my experience it does not work,
                                                    > > > not exactly.
                                                    > > >
                                                    > > > None of this automatic stuff really works very well above the bass. But this
                                                    > > > is not to say that SOMETHING does not work well above the bass. As I used to
                                                    > > > say to GF, since you can change the sound in the higher frequencies you are
                                                    > > > almost bound to be able to make it better!
                                                    > > >
                                                    > > > The problem is the same as audio has always had. Engineers like engineering
                                                    > > > solutions, and solutions that involve human experience as an evaluation tool
                                                    > > > make engineers nervous, in general.
                                                    > > >
                                                    > > > I think it makes a whole lot of sense simply to get pink noise to sound
                                                    > > > right. Of course you have somehow to find out what right means, as TM
                                                    > > > suggested. But you can experiment and learn to hear what is right quite
                                                    > > > easily. Just try a lot of EQs both on pink noise and on recordings that you
                                                    > > > think ought to be good and set up something that makes them sound right in
                                                    > > > some intuitive sense.
                                                    > > >
                                                    > > > This could take some time! But it beats messing around trying to do the same
                                                    > > > thing by buying different equipment items or modifying your room. And it is
                                                    > > > a lot cheaper too.
                                                    > > >
                                                    > > > Of course one needs some sort of half decent listening room to begin with.
                                                    > > > But that is not so hard to arrange.
                                                    > > >
                                                    > > > I think the big failing here is human--people have not worked with user
                                                    > > > controlled EQ carefully enough. Talk about laziness! People would apparently
                                                    > > > rather buy stuff than work even a little.
                                                    > > > This is craziness to my mind. Anyway, once you get into it, it is FUN
                                                    > > > to play with EQ.
                                                    > > >
                                                    > > > I find it really intriguing to try to adjust one speaker to sound like
                                                    > > > another. Interesting to try!
                                                    > > >
                                                    > > > And don't forget that NO speaker, essentially , is as smooth as it ought to
                                                    > > > be. Just look at those NRC measurements. Forget questions of what radiation
                                                    > > > pattern ought to be for the moment and just look at one axis response.
                                                    > > > ALmost NO speaker is as smooth as it ought to be.
                                                    > > > It very seldom happens. And when adds in the effects of off axis--well, I
                                                    > > > think it is fair to say that it NEVER happens, that it is as smooth as it
                                                    > > > ought to be(independently of what "flat " means, everyone agrees that things
                                                    > > > ought to be really smooth--but they are not, not in detail. Even getting
                                                    > > > totally smooth 1/3 octave is quite unusual without EQ).
                                                    > > >
                                                    > > > Anyway, anything with a name like Autocal is probably not really a total
                                                    > > > soultion above the bass --because if it is Auto it is going to produce
                                                    > > > almost surely the WRONG cal(ibration), the right one being a matter of
                                                    > > > listening.(I did not watch long enough to find out whether Autocal works
                                                    > > > above the bass--the name is just an example).
                                                    > > >
                                                    > > > Engineers don't like this idea and most of them do not trust the customers
                                                    > > > either. But it is your music--and your head! whose transfer function you
                                                    > > > listen to.
                                                    > > >
                                                    > > > REG
                                                    > > >
                                                    > > >
                                                    > > >
                                                    > > >
                                                    > > > ------------------------------------
                                                    > > >
                                                    > > > Yahoo! Groups Links
                                                    > > >
                                                    > > >
                                                    > > >
                                                    > > >
                                                    > > >
                                                    > > > ------------------------------------
                                                    > > >
                                                    > > > Yahoo! Groups Links
                                                    > > >
                                                    > >
                                                    > >
                                                    > >
                                                    > >
                                                    > > ------------------------------------
                                                    > >
                                                    > > Yahoo! Groups Links
                                                    > >
                                                    >
                                                    >
                                                    >
                                                    >
                                                    > ------------------------------------
                                                    >
                                                    > Yahoo! Groups Links
                                                    >




                                                    No virus found in this incoming message.
                                                    Checked by AVG - www.avg.com
                                                    Version: 9.0.862 / Virus Database: 271.1.1/3189 - Release Date: 10/10/10 14:34:00
                                                  Your message has been successfully submitted and would be delivered to recipients shortly.