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

Re: [micbuilders] Re: why not per microphone equalization ?

Expand Messages
  • Jerry Lee Marcel
    ... Yes. That s the case of any cardioid or figure-8 mic ... It is sufficient that it happens at any frequency to make the whole system non-minimum-phase. ...
    Message 1 of 51 , Jul 2, 2011
    • 0 Attachment
      Le 02/07/2011 14:13, Richard Lee a écrit :
      >
      > I can only comment briefly ..
      >
      > There are a number of necessary but not sufficient conditions for
      > non-Minimum Phase behaviour. Two of these are
      >
      > 1 There must be more than 1 path for the signal
      >
      Yes. That's the case of any cardioid or figure-8 mic
      >
      > 2 The original arrival must be smaller than later arrivals
      >
      It is sufficient that it happens at any frequency to make the whole
      system non-minimum-phase.
      >
      >
      > It so happens that hardly any microphones are non-minimum phase. It would
      > be a very poor microphone that was.
      >
      Most SDC's can be considered min-phase in the audio spectrum; LDC's
      maybe not.
      >
      >
      > One high-quality microphone that might be, is a tetrahedral soundfield
      > mike, as multiple capsules contribute to each output. I went to a lot of
      > trouble circa 1980 to check this and was pleased to find that the Calrec
      > Soundfield was Minimum Phase cos I could only apply Analogue Minimum
      > Phase
      > EQ at that time. Measuring Acoustic Phase was MUCH more difficult in dem
      > days.
      >
      > I have been able to confirm this in the 21st century.
      >
      > On these, correcting the complex frequency response also corrects the
      > impulse response.
      > _________________________
      >
      > > Specifically, I want a physical/mathematical proof that by taking a
      > quality measurement microphone and by applying on a recording made
      > with it
      > complex (i.e. including phase) frequency plot of U47 I won't hear the
      > same
      > sound as hear from U47 itself.
      >
      > Err..rh You are asking for mathematical proof of a psycho-acoustic
      > (subjective) phenomena. The onus is in fact on you to show the converse.
      > ie
      >
      > Can you dream up a complex EQ which when applied to a number of
      > recordings
      > made with some specified microphone, fool a number of U47 diehards that
      > these recordings were made with their favourite mike in Double Blind
      > Listening tests bla bla.
      >
      > I actually think it is possible but to make it easier, I suggest you
      > choose
      > a double diaphragm mike about the same size. You'll probably get away
      > with
      > a AKG 414, Debenham, Robinson & Stebbings or Rode NT2. You might have to
      > EQ back & front separately.
      >
      > All dis FFT, cepstrum stuff assumes the system is linear time invariant.
      > Most of the time, this assumption is useful but microphones and
      > particularly speakers, are definitely NOT linear time invariant.
      > _________________________
      >
      > There are many caveats with Digital EQ. Foremost is that your
      > measurements
      > have artifacts & noise. Careful smoothing is necessary to ameliorate some
      > of these without losing info. In Digital EQ, what you DON'T EQ is as
      > (maybe more) important as what you do. One related example is head basket
      > reflections which are probably best NOT EQ'd out.
      >
      >


      [Non-text portions of this message have been removed]
    • Jerry Lee Marcel
      That s what I thought; it s not really looking at ... ... [Non-text portions of this message have been removed]
      Message 51 of 51 , Jul 7, 2011
      • 0 Attachment
        That's what I thought; it's not really "looking at"...

        Le 07/07/2011 17:20, Richard Lee a écrit :
        >
        > > I would be interested to know how, from looking at an impulse
        > response, one can infer it is minimum-phase or not.
        >
        > The test for Minimum Phase is to compare the phase response with the
        > Hilbert Transform of the Amplitude Response. see eg
        >
        > "Digital Signal Processing" - Oppenheim & Schafer
        > "Is Linear Phase Worthwhile?" - Lee, AES Hamburg, 1981(?)
        >
        > With an Impulse Response eg as a *.WAV file, I ..
        >
        > FFT to get the complex frequency resp.
        > Natural Log to get Amplitude & Phase
        > Digital Hilbert Transform on the Log Amplitude
        > Compare with the Phase.
        >
        > I've been doing Analogue & Digital EQ for speakers since 1977 with
        > some success. This millenium, it's been mostly mikes.
        >
        > More on the "linear time invariant assumption" and its caveats later ...
        >
        > Also Eric has the right idea. Speakers & mikes are 1D to 3D converters
        > & vice versa.
        >
        > This is part of the problem, " What you Do or Don't EQ ". Your digital
        > or analogue EQ is only a 1D processor.
        >
        >


        [Non-text portions of this message have been removed]
      Your message has been successfully submitted and would be delivered to recipients shortly.