I have been trying to develop some microphone testing software that measure frequency response of microphone. It should be possible using the impulse response or FFT of recieved waves of the microphone. Could anybody explain me how to ?? Any help would be much appreciated. Many thanks Robert
Frequency response
Started by ●February 2, 2010
Reply by ●February 2, 20102010-02-02
On 02/02/2010 1:34 PM, PeteK wrote:> I have been trying to develop some microphone testing software that measure > frequency response of microphone. > > It should be possible using the impulse response or FFT of recieved waves > of > the microphone. > > Could anybody explain me how to ?? > > Any help would be much appreciated. > > Many thanks > > Robert > > > > > >To properly test a microphone, you should also use another microphone with a flat frequency response suitable for use as a reference: [1] http://www.dpamicrophones.com/en/Microphone-University/Technology-Guide/How%20to%20test%20the%20performance%20of%20a%20microphone.aspx [2] http://personal.ee.surrey.ac.uk/Personal/P.Jackson/ee2.lab/XX_micfr/ Both microphones can be attached to separate channels of an audio ADC. IMHO, one of the "standard" methods to test a microphone is the reciprocal method: [3] http://ts.nist.gov/MeasurementServices/Calibrations/upload/EAcous.pdf In [2] above, white noise or a sinusoidal signal is mentioned as being suitable as a test source. You could probably also use frequency-sweeps, chirps or Maximum Length Sequences. Frequency responses can be obtained via FFT or other spectral estimation techniques (i.e. Yule-Walker AR, MUSIC). You don't require a large lab setup to do the test, but perhaps the best thing to do would be to use more than one microphone.
Reply by ●February 3, 20102010-02-03
PeteK wrote:> I have been trying to develop some microphone testing software that measure > frequency response of microphone. > > It should be possible using the impulse response or FFT of recieved waves > of > the microphone. > > Could anybody explain me how to ?? > > Any help would be much appreciated. > > Many thanks > > Robert > > > > > >One technique - best done outside - is to turn a thift-store camera flash into a spark-gap generator. You record two sample streams into two channels - one is from a simple loop pickup around the wire of the spark gap generator, the other is the microphone signal. Deconvolve the mic. signal against the inductive pickup signal, and you get the impulse signature of the mic. Voxengo deconvolver is free. And be careful - very high V in a camera flash. -- Les Cargill
Reply by ●February 3, 20102010-02-03
On 2 Feb, 23:37, Nicholas Kinar <n.ki...@usask.ca> wrote:> Frequency responses can be obtained via FFT or other spectral estimation > techniques (i.e. Yule-Walker AR, MUSIC).Do *only* use the DFT or DFT-based non-parmetric methods for these kinds of measurements!! The parametric methods are riddled with numeric artifacts to the extent that any results will be dominated by the quirks and idiosyncracies of the method rather than the measured data. Rune
Reply by ●February 3, 20102010-02-03
Les Cargill wrote:> PeteK wrote: >> I have been trying to develop some microphone testing software that >> measure >> frequency response of microphone. >> >> It should be possible using the impulse response or FFT of recieved waves >> of >> the microphone. >> >> Could anybody explain me how to ?? >> >> Any help would be much appreciated. >> >> Many thanks >> >> Robert >> >> >> >> >> >> > > > One technique - best done outside - is to turn a thift-store > camera flash into a spark-gap generator. You record two sample streams > into two channels - one is from a simple loop pickup around the wire > of the spark gap generator, the other is the microphone signal. > > Deconvolve the mic. signal against the inductive pickup signal, > and you get the impulse signature of the mic. > > Voxengo deconvolver is free. And be careful - very high V in a > camera flash.Best done outside in the midst of a field of grain or tall grass. According to many acousticians (including Paul Klipsh and George Briggs of Wharfdale) that environment is very anechoic. Thanks, Les, for the lead to Voxengo. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●February 3, 20102010-02-03
Thank you for your respons, I will explore your suggestion. That is I was looking... Robert>On 02/02/2010 1:34 PM, PeteK wrote: >> I have been trying to develop some microphone testing software thatmeasure>> frequency response of microphone. >> >> It should be possible using the impulse response or FFT of recievedwaves>> of >> the microphone. >> >> Could anybody explain me how to ?? >> >> Any help would be much appreciated. >> >> Many thanks >> >> Robert >> >> >> >> >> >> > >To properly test a microphone, you should also use another microphone >with a flat frequency response suitable for use as a reference: > >[1] >http://www.dpamicrophones.com/en/Microphone-University/Technology-Guide/How%20to%20test%20the%20performance%20of%20a%20microphone.aspx > >[2] http://personal.ee.surrey.ac.uk/Personal/P.Jackson/ee2.lab/XX_micfr/ > >Both microphones can be attached to separate channels of an audio ADC. > >IMHO, one of the "standard" methods to test a microphone is the >reciprocal method: > >[3]http://ts.nist.gov/MeasurementServices/Calibrations/upload/EAcous.pdf> >In [2] above, white noise or a sinusoidal signal is mentioned as being >suitable as a test source. You could probably also use >frequency-sweeps, chirps or Maximum Length Sequences. > >Frequency responses can be obtained via FFT or other spectral estimation>techniques (i.e. Yule-Walker AR, MUSIC). > >You don't require a large lab setup to do the test, but perhaps the best>thing to do would be to use more than one microphone. > >
Reply by ●February 3, 20102010-02-03
On Feb 3, 11:58�am, Jerry Avins <j...@ieee.org> wrote:> Les Cargill wrote: > > PeteK wrote: > >> I have been trying to develop some microphone testing software that > >> measure > >> frequency response of microphone. > > >> It should be possible using the impulse response or FFT of recieved waves > >> of > >> the microphone. > > >> Could anybody explain me how to ?? > > >> Any help would be much appreciated. > > >> Many thanks > > >> Robert > > > One technique - best done outside - is to turn a thift-store > > camera flash into a spark-gap generator. You record two sample streams > > into two channels - one is from a simple loop pickup around the wire > > of the spark gap generator, the other is the microphone signal. > > > Deconvolve the mic. signal against the inductive pickup signal, > > and you get the impulse signature of the mic. > > > Voxengo deconvolver is free. And be careful - very high V in a > > camera flash. > > Best done outside in the midst of a field of grain or tall grass. > According to many acousticians (including Paul Klipsh and George Briggs > of Wharfdale) that environment is very anechoic. > > Thanks, Les, for the lead to Voxengo. > > Jerry > -- > Engineering is the art of making what you want from things you can get. > �����������������������������������������������������������������������- Hide quoted text - > > - Show quoted text -well the whole point of using the impulse response method as opposed to swept sines is that you can time window the impulse response to remove the effect of the environment... Mark
Reply by ●February 3, 20102010-02-03
Mark wrote:> On Feb 3, 11:58 am, Jerry Avins <j...@ieee.org> wrote: >> Les Cargill wrote: >>> PeteK wrote: >>>> I have been trying to develop some microphone testing software that >>>> measure >>>> frequency response of microphone. >>>> It should be possible using the impulse response or FFT of recieved waves >>>> of >>>> the microphone. >>>> Could anybody explain me how to ?? >>>> Any help would be much appreciated. >>>> Many thanks >>>> Robert >>> One technique - best done outside - is to turn a thift-store >>> camera flash into a spark-gap generator. You record two sample streams >>> into two channels - one is from a simple loop pickup around the wire >>> of the spark gap generator, the other is the microphone signal. >>> Deconvolve the mic. signal against the inductive pickup signal, >>> and you get the impulse signature of the mic. >>> Voxengo deconvolver is free. And be careful - very high V in a >>> camera flash. >> Best done outside in the midst of a field of grain or tall grass. >> According to many acousticians (including Paul Klipsh and George Briggs >> of Wharfdale) that environment is very anechoic. >> >> Thanks, Les, for the lead to Voxengo. >> >> Jerry >> -- >> Engineering is the art of making what you want from things you can get. >> �����������������������������������������������������������������������- Hide quoted text - >> >> - Show quoted text - > > well the whole point of using the impulse response method as opposed > to swept sines is that you can time window the impulse response to > remove the effect of the environment...That the impulse response is short doesn't guarantee that it's negligibly short. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●February 3, 20102010-02-03
Mark wrote:> On Feb 3, 11:58 am, Jerry Avins <j...@ieee.org> wrote: >> Les Cargill wrote: >>> PeteK wrote: >>>> I have been trying to develop some microphone testing software that >>>> measure >>>> frequency response of microphone. >>>> It should be possible using the impulse response or FFT of recieved waves >>>> of >>>> the microphone. >>>> Could anybody explain me how to ?? >>>> Any help would be much appreciated. >>>> Many thanks >>>> Robert >>> One technique - best done outside - is to turn a thift-store >>> camera flash into a spark-gap generator. You record two sample streams >>> into two channels - one is from a simple loop pickup around the wire >>> of the spark gap generator, the other is the microphone signal. >>> Deconvolve the mic. signal against the inductive pickup signal, >>> and you get the impulse signature of the mic. >>> Voxengo deconvolver is free. And be careful - very high V in a >>> camera flash. >> Best done outside in the midst of a field of grain or tall grass. >> According to many acousticians (including Paul Klipsh and George Briggs >> of Wharfdale) that environment is very anechoic. >> >> Thanks, Les, for the lead to Voxengo. >> >> Jerry >> -- >> Engineering is the art of making what you want from things you can get. >> �����������������������������������������������������������������������- Hide quoted text - >> >> - Show quoted text - > > well the whole point of using the impulse response method as opposed > to swept sines is that you can time window the impulse response to > remove the effect of the environment... > > Mark >A mic/preamp combo's impulse signature will be very short. -- Les Cargill
Reply by ●February 3, 20102010-02-03
Jerry Avins wrote:> Les Cargill wrote: >> PeteK wrote: >>> I have been trying to develop some microphone testing software that >>> measure >>> frequency response of microphone. >>> >>> It should be possible using the impulse response or FFT of recieved >>> waves >>> of >>> the microphone. >>> >>> Could anybody explain me how to ?? >>> >>> Any help would be much appreciated. >>> >>> Many thanks >>> >>> Robert >>> >>> >>> >>> >>> >>> >> >> >> One technique - best done outside - is to turn a thift-store >> camera flash into a spark-gap generator. You record two sample streams >> into two channels - one is from a simple loop pickup around the wire >> of the spark gap generator, the other is the microphone signal. >> >> Deconvolve the mic. signal against the inductive pickup signal, >> and you get the impulse signature of the mic. >> >> Voxengo deconvolver is free. And be careful - very high V in a >> camera flash. > > Best done outside in the midst of a field of grain or tall grass. > According to many acousticians (including Paul Klipsh and George Briggs > of Wharfdale) that environment is very anechoic. >Given the weather lately, I imagine a snowy field would be perfect.> Thanks, Les, for the lead to Voxengo. >You are most welcome. Very handy to have around.> Jerry-- Les Cargill
