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

Using Spectrum Lab for sub audio

Expand Messages
  • ayepschryme
    Hello Wolf and all, Thank you for allowing me on this group, I am new here. Also a big thank you to you Wolf for an excellent peice of software. I am
    Message 1 of 9 , Mar 8, 2011
    • 0 Attachment
      Hello Wolf and all, Thank you for allowing me on this group, I am new here. Also a big thank you to you Wolf for an excellent peice of software. I am interested in making infrasound measurements as part of a research group studying among other things, health effects of low frequency noise generated by alternative energy wind turbines. First I noticed that my web cam is sensitive below 20Hz quite by accident while using spectrum lab to measure the resonance of an engine intake airbox. When I set the frequency limits between zero and 20Hz though, I notice the frequency resolution displayed goes to roughly 1Hz blocks so I am wondering if anyone can advise me how to get the best resolution settings to give a good compromise between time and frequency resolution. I tried changing a few parameters but there are a lot of them and I don't really understand it well.

      It seems most people doing extremely low sound measurements have to build their own microphone and preamp and I have found some circuit examples but all analog. Since there are some very easy to use USB to serial or parallel solutions these days I thought that might be a better option for input to a laptop. Anyone care to comment on that as an option for using Spectrum Lab?

      72
      Joe, ve3vxo
    • sm5le
      Hi Joe I am also measuring infrasound. Using: SpecLab Loudspeaker as microphone. EMU 0202 as A/D
      Message 2 of 9 , Mar 8, 2011
      • 0 Attachment
        Hi Joe
        I am also measuring infrasound. Using:
        SpecLab
        Loudspeaker as microphone.
        EMU 0202 as A/D ;http://www.emu.com/products/product.asp?category=610&subcategory=611&product=15186
        Also tryed INA217 as preeamp.
        I have aslo "probelm" with finding best adjustments (if there is one :-) but this helps a lot : http://www.vlf.it/fft_beginners/fft_beginners.html

        73 Sven SM5LE

        --- In SpectrumLabUsers@yahoogroups.com, "ayepschryme" <jstreet@...> wrote:
        >
        > Hello Wolf and all, Thank you for allowing me on this group, I am new here. Also a big thank you to you Wolf for an excellent peice of software. I am interested in making infrasound measurements as part of a research group studying among other things, health effects of low frequency noise generated by alternative energy wind turbines. First I noticed that my web cam is sensitive below 20Hz quite by accident while using spectrum lab to measure the resonance of an engine intake airbox. When I set the frequency limits between zero and 20Hz though, I notice the frequency resolution displayed goes to roughly 1Hz blocks so I am wondering if anyone can advise me how to get the best resolution settings to give a good compromise between time and frequency resolution. I tried changing a few parameters but there are a lot of them and I don't really understand it well.
        >
        > It seems most people doing extremely low sound measurements have to build their own microphone and preamp and I have found some circuit examples but all analog. Since there are some very easy to use USB to serial or parallel solutions these days I thought that might be a better option for input to a laptop. Anyone care to comment on that as an option for using Spectrum Lab?
        >
        > 72
        > Joe, ve3vxo
        >
      • ayepschryme
        Hi Sven; Thanks for your info. But how can you measure infrasound using EMU 0202 as A/D?? The specs give 20hz as lower corner frequency. The best
        Message 3 of 9 , Mar 8, 2011
        • 0 Attachment
          Hi Sven;

          Thanks for your info. But how can you measure infrasound using EMU 0202 as A/D?? The specs give 20hz as lower corner frequency. The best 'microphone I have found is the honeywell differential pressure transducer which is modified with a capillary leak on one port to tune the bandwidth of the response or one port can be blocked when measuring for short duration under stable ambient pressure. The response is flat down to DC. You can find information on how to build this sensor here:


          http://www.ees.nmt.edu/johnson_j/MICROPHONES/microphone_list.html

          Best regards
          Joe



          --- In SpectrumLabUsers@yahoogroups.com, "sm5le" <sm5le@...> wrote:
          >
          > Hi Joe
          > I am also measuring infrasound. Using:
          > SpecLab
          > Loudspeaker as microphone.
          > EMU 0202 as A/D ;http://www.emu.com/products/product.asp?category=610&subcategory=611&product=15186
          > Also tryed INA217 as preeamp.
          > I have aslo "probelm" with finding best adjustments (if there is one :-) but this helps a lot : http://www.vlf.it/fft_beginners/fft_beginners.html
          >
          > 73 Sven SM5LE
          >
        • wolf_dl4yhf
          Hello Sven, Joe, and group, Spectrum Lab itself doesn t care about the frequency range. For infrasound, use a low sampling rate, a long FFT, and (if that still
          Message 4 of 9 , Mar 8, 2011
          • 0 Attachment
            Hello Sven, Joe, and group,

            Spectrum Lab itself doesn't care about the frequency range. For infrasound, use a low sampling rate, a long FFT, and (if that still doesn't give enough frequency resolution) decimation to reduce the sampling rate even further.
            Some soundcards (not sure about the E-MU) only have a single coupling capacitor in the signal path; removing it gives a flat response down to DC. I have successfully modified the internal audio chip in an old notebook PC, where the capacitor was easy to find.
            An alternative to get DC capability from an ordinary soundcard would be to feed the signal into a voltage-to-frequency converter, and demodulate the FM signal with spectrum lab.

            All the best,
               Wolf .

            Am 08.03.2011 18:38, schrieb ayepschryme:
             

            Hi Sven;

            Thanks for your info. But how can you measure infrasound using EMU 0202 as A/D?? The specs give 20hz as lower corner frequency. The best 'microphone I have found is the honeywell differential pressure transducer which is modified with a capillary leak on one port to tune the bandwidth of the response or one port can be blocked when measuring for short duration under stable ambient pressure. The response is flat down to DC. You can find information on how to build this sensor here:

            http://www.ees.nmt.edu/johnson_j/MICROPHONES/microphone_list.html

            Best regards
            Joe

            --- In SpectrumLabUsers@yahoogroups.com, "sm5le" <sm5le@...> wrote:
            >
            > Hi Joe
            > I am also measuring infrasound. Using:
            > SpecLab
            > Loudspeaker as microphone.
            > EMU 0202 as A/D ;http://www.emu.com/products/product.asp?category=610&subcategory=611&product=15186
            > Also tryed INA217 as preeamp.
            > I have aslo "probelm" with finding best adjustments (if there is one :-) but this helps a lot : http://www.vlf.it/fft_beginners/fft_beginners.html
            >
            > 73 Sven SM5LE
            >


          • sm5le
            Joe Me just happened to have EMU0202 (from music recordings) and tested response down and finde abt 2-3 Hz for 3 dB corner. BUT the problem is not freq.
            Message 5 of 9 , Mar 8, 2011
            • 0 Attachment
              Joe
              Me just happened to have EMU0202 (from music recordings) and tested response down and finde abt 2-3 Hz for 3 dB corner. BUT the problem is not freq. response it is the 1/f noise (that noise increasing with lowering freq.) and EMU just happened to be good in that case. I have not seen a circuit diagram for EMU0202 so I do not know abt coupling capacitors etc.
              The guitar cube speaker (such box with amplifier and loudspeaker in same box) works OK down ~3Hz as well with pree amp (also 1/f problem).
              Hope that helps.
              /Sven
              ps OK Wolf abt sampling etc.


              --- In SpectrumLabUsers@yahoogroups.com, "ayepschryme" <jstreet@...> wrote:
              >
              > Hi Sven;
              >
              > Thanks for your info. But how can you measure infrasound using EMU 0202 as A/D?? The specs give 20hz as lower corner frequency. The best 'microphone I have found is the honeywell differential pressure transducer which is modified with a capillary leak on one port to tune the bandwidth of the response or one port can be blocked when measuring for short duration under stable ambient pressure. The response is flat down to DC. You can find information on how to build this sensor here:
              >
              >
              > http://www.ees.nmt.edu/johnson_j/MICROPHONES/microphone_list.html
              >
              > Best regards
              > Joe
              >
              >
              >
              > --- In SpectrumLabUsers@yahoogroups.com, "sm5le" <sm5le@> wrote:
              > >
              > > Hi Joe
              > > I am also measuring infrasound. Using:
              > > SpecLab
              > > Loudspeaker as microphone.
              > > EMU 0202 as A/D ;http://www.emu.com/products/product.asp?category=610&subcategory=611&product=15186
              > > Also tryed INA217 as preeamp.
              > > I have aslo "probelm" with finding best adjustments (if there is one :-) but this helps a lot : http://www.vlf.it/fft_beginners/fft_beginners.html
              > >
              > > 73 Sven SM5LE
              > >
              >
            • Joe Street
              Hi Wolf; The long FFT does increase the resolution but also slows down the waterfall update. A lower sampling rate does the same. So I want the best of both
              Message 6 of 9 , Mar 8, 2011
              • 0 Attachment
                Hi Wolf;

                The long FFT does increase the resolution but also slows down the waterfall update.  A lower sampling rate does the same.  So I want the best of both worlds.  I tried various settings and found with a high sampling rate and long FFT I get reasonable refresh rate and still about 1 hz resolution which is useable. I am using a lenovo laptop with about 2Ghz and 2Gb RAM.  I suspect the internal mic and soundcard rolls off somewhere around 5 to 8hz. I don't want to open it because it belongs to my employer but I am willing to open the webcam I have and see what is there in terms of coupling capacitor and what I can do.  In that case I may have less control of the sampling rate but I'll see.

                I could do something with a small PIC and a prolific or FTDI serial to USB chip.  The drivers for these are handled just like a COM port by the OS so do you think it would be easy to get it talking to SL?

                Thank you for your help.

                Joe





                On 08/03/2011 1:04 PM, wolf_dl4yhf wrote:
                 

                Hello Sven, Joe, and group,

                Spectrum Lab itself doesn't care about the frequency range. For infrasound, use a low sampling rate, a long FFT, and (if that still doesn't give enough frequency resolution) decimation to reduce the sampling rate even further.
                Some soundcards (not sure about the E-MU) only have a single coupling capacitor in the signal path; removing it gives a flat response down to DC. I have successfully modified the internal audio chip in an old notebook PC, where the capacitor was easy to find.
                An alternative to get DC capability from an ordinary soundcard would be to feed the signal into a voltage-to-frequency converter, and demodulate the FM signal with spectrum lab.

                All the best,
                   Wolf .


              • wolf_dl4yhf
                Hi Joe, ... Yes, that s one option. And if you use the same data format as I was using in my own PIC-based A/D converter, you don t need to write any special
                Message 7 of 9 , Mar 8, 2011
                • 0 Attachment
                  Hi Joe,

                  you wrote:


                   

                  Hi Wolf;

                  The long FFT does increase the resolution but also slows down the waterfall update.  A lower sampling rate does the same.  So I want the best of both worlds.  I tried various settings and found with a high sampling rate and long FFT I get reasonable refresh rate and still about 1 hz resolution which is useable. I am using a lenovo laptop with about 2Ghz and 2Gb RAM.  I suspect the internal mic and soundcard rolls off somewhere around 5 to 8hz. I don't want to open it because it belongs to my employer but I am willing to open the webcam I have and see what is there in terms of coupling capacitor and what I can do.  In that case I may have less control of the sampling rate but I'll see.

                  I could do something with a small PIC and a prolific or FTDI serial to USB chip.  The drivers for these are handled just like a COM port by the OS so do you think it would be easy to get it talking to SL?

                  Yes, that's one option. And if you use the same data format as I was using in my own PIC-based A/D converter, you don't need to write any special driver for it:

                  http://www.qsl.net/dl4yhf/soundutl/serpicad.htm


                  Another alternative (which I'd prefer these days) is a homebrew soundcard as described in a german radio amateur magazine... guess they were using a PCM2900 (just google for it). I didn't check the TI datasheet yet, but this chip may even go down to DC with suitable input circuitry. The big advantage is that you don't need to write any driver software for it.

                  Cheers,
                     Wolf .

                • Jan Sjöholm
                  Re: Voltage to frequency Hi Wolf, I have thought of using a VF converter for sub audio, but never tested it with SL. Thanks for the encouragement. I am
                  Message 8 of 9 , Apr 17, 2011
                  • 0 Attachment
                    Re: Voltage to frequency
                    Hi Wolf,
                    I have thought of using a VF converter for sub audio, but never tested it with SL. Thanks for the encouragement. I am familiar with the LM331, it will do the job well.
                    Best regards
                    Jan


                    --- In SpectrumLabUsers@yahoogroups.com, wolf_dl4yhf <dl4yhf@...> wrote:
                    >
                    > Hello Sven, Joe, and group,
                    >
                    > Spectrum Lab itself doesn't care about the frequency range. For
                    > infrasound, use a low sampling rate, a long FFT, and (if that still
                    > doesn't give enough frequency resolution) decimation to reduce the
                    > sampling rate even further.
                    > Some soundcards (not sure about the E-MU) only have a single coupling
                    > capacitor in the signal path; removing it gives a flat response down to
                    > DC. I have successfully modified the internal audio chip in an old
                    > notebook PC, where the capacitor was easy to find.
                    > An alternative to get DC capability from an ordinary soundcard would be
                    > to feed the signal into a voltage-to-frequency converter, and demodulate
                    > the FM signal with spectrum lab.
                    >
                    > All the best,
                    > Wolf .
                    >
                    > Am 08.03.2011 18:38, schrieb ayepschryme:
                    > >
                    > > Hi Sven;
                    > >
                    > > Thanks for your info. But how can you measure infrasound using EMU
                    > > 0202 as A/D?? The specs give 20hz as lower corner frequency. The best
                    > > 'microphone I have found is the honeywell differential pressure
                    > > transducer which is modified with a capillary leak on one port to tune
                    > > the bandwidth of the response or one port can be blocked when
                    > > measuring for short duration under stable ambient pressure. The
                    > > response is flat down to DC. You can find information on how to build
                    > > this sensor here:
                    > >
                    > > http://www.ees.nmt.edu/johnson_j/MICROPHONES/microphone_list.html
                    > >
                    > > Best regards
                    > > Joe
                    > >
                    > > --- In SpectrumLabUsers@yahoogroups.com
                    > > <mailto:SpectrumLabUsers%40yahoogroups.com>, "sm5le" <sm5le@> wrote:
                    > > >
                    > > > Hi Joe
                    > > > I am also measuring infrasound. Using:
                    > > > SpecLab
                    > > > Loudspeaker as microphone.
                    > > > EMU 0202 as A/D
                    > > ;http://www.emu.com/products/product.asp?category=610&subcategory=611&product=15186
                    > > <http://www.emu.com/products/product.asp?category=610&subcategory=611&product=15186>
                    > > > Also tryed INA217 as preeamp.
                    > > > I have aslo "probelm" with finding best adjustments (if there is one
                    > > :-) but this helps a lot :
                    > > http://www.vlf.it/fft_beginners/fft_beginners.html
                    > > >
                    > > > 73 Sven SM5LE
                    > > >
                    > >
                    > >
                    >
                  • yewie56
                    Hi, on www.brummforum.net we are investigating some effects of the so called Taos-Hum Phenomena and are creating SpectrumLab waterfalls. Sorry, corespondance
                    Message 9 of 9 , May 5, 2011
                    • 0 Attachment
                      Hi,

                      on www.brummforum.net we are investigating some effects of the so called "Taos-Hum Phenomena" and are creating SpectrumLab waterfalls. Sorry, corespondance language is german, but english should be welcome too.

                      I think, you can use any microphone (nearly, not all!), if you can calibrate the frequency response, but be aware of the 1/f noise.
                      We are using some simple Desktop-Microphones with a good subsonic response:

                      See the third response curve from above ( Elektret (78) )

                      http://audioclub.de/index.php?option=com_content&view=article&id=367:500-mikrofonkalibrierungen-eine-uebersicht&catid=36:software--messtechnik&Itemid=66

                      The response below 10Hz is affordable. One interesting feature is, that these inexpensive microphones can be calibrated.

                      A calibrated one you can buy here for 25 Euros.

                      http://lasip.hifi-selbstbau.de/index.php?page=shop.product_details&flypage=flypage-ask.tpl&product_id=17&category_id=1&option=com_virtuemart&Itemid=64&vmcchk=1&Itemid=64

                      Better choice:
                      Or buy a 2Euro Microphone DIGITUS Multimedia Microphone DA-10101,
                      and send it together with your USB-Soundcard to the 25 Euro calibration service

                      http://lasip.hifi-selbstbau.de/index.php?page=shop.product_details&flypage=flypage.tpl&product_id=19&category_id=9&option=com_virtuemart&Itemid=64

                      You will get a frequency response deviation file. (and your hardware back)

                      Im investigating, if SpectrumLab can handle this calibration file to get a plain linear subsonic frequency response.

                      yewie56
                    Your message has been successfully submitted and would be delivered to recipients shortly.