Subject:about physics and mathematics...From:"Antonio S. Pena" <apena(at)TSC.UVIGO.ES>Date:Thu, 24 Jun 1999 13:02:05 +0200--------------28A7DFE5701F47FE2717839C Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > > In Praat, the maximum absolute value of the samples is 1, which, if considered > > Pascals, corresponds to an intensity of 94 dB re auditory threshold. > > [...] you should use a dB meter once, > > with your computer set to a maximum input gain, and use the same gain > > setting for all subsequent recordings. > > > Date: Wed, 23 Jun 1999 21:42:27 -0600 From: Paul Boersma <boersma(at)FON.HUM.UVA.NL> Subject: Re: cepstrum: computing epsilon > On Jun 23, 7:46pm, Laszlo Toth wrote: > > > FFT (or, rather, DFT) is a formula that > > converts a series of numbers into another series of numbers. For me. > > (...) Now I see that by > > "Praat fft" you meant "the spectrum analysis facilities (and the belonging > > objects) in the Praat software package". > > Exactly. That is the advantage of having a program that models the > physics (like Praat) instead of the mathematics (like Matlab). > > Someone complained about the problem of choosing the "epsilon" in > a Matlab implementation of the cepstrum. In a Praat script, you > can express this epsilon in physical units. So if you want to add > white noise with a density of 10^-8 Pascal/Hertz, your cepstrum > routine would be (...) hi list, just a brief comment about maths and physics... the advantage of having a program that deals directly with Physics units (as Praat's) is that the calibration is already done. If working with matlab, the mathematical models that render the physic behaviour must be proper calibrated to normalize the dynamic range (p.e: +32767/-32768 if your input is 16 bits quantifed) to a given value in a physic unit, say, for example, SPL units. If you feed a full-range 16 bit signal into a soundcard, output SPL depends ONLY on the gain applied to the amplifier that attacks the transducer (headph or loudspk). So, you can have a, let's say, 50 dB SPL signal with a +/- 32767 sine signal or with a +/- 16383 sine signal if the amplifier gain is modified properly. The difference between both cases is that the background noise increases but SPL pressure stands the same. The problem of calibration is a very sensitive problem that is not always well managed and results coming from a incorrect calibration will probably give a wrong, even astonishing, result. But please remember that physics modelling relies on a, let's say, 40%?empyrical data and a 60%? numerical modelling, so a math environment is well suited for physics models ONLY IF we use the correct *reality->virtuality* or *virtuality->reality* mapping. Hope it helps and thank both to Praat and Matlab developers for providing us with such useful tools have a nice day you all Antonio -- Antonio S. Pena E.T.S.E.Telecomunicacion Universidade de Vigo 36200 - VIGO ( Spain ) Phone: +34-986-812-130 FAX: +34-986-812-116 E-mail: apena(at)tsc.uvigo.es --------------28A7DFE5701F47FE2717839C Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML> <BLOCKQUOTE TYPE=CITE>> In Praat, the maximum absolute value of the samples is 1, which, if considered <BR>> Pascals, corresponds to an intensity of 94 dB re auditory threshold. <BR>> [...] you should use a dB meter once, <BR>> with your computer set to a maximum input gain, and use the same gain <BR>> setting for all subsequent recordings. <BR>> <BR> </BLOCKQUOTE> <P><BR>Date: Wed, 23 Jun 1999 21:42:27 -0600 <BR>From: Paul Boersma <boersma(at)FON.HUM.UVA.NL> <BR>Subject: Re: cepstrum: computing epsilon <BLOCKQUOTE TYPE=CITE>On Jun 23, 7:46pm, Laszlo Toth wrote: <P>> FFT (or, rather, DFT) is a formula that <BR>> converts a series of numbers into another series of numbers. For me. <BR>> (...) Now I see that by <BR>> "Praat fft" you meant "the spectrum analysis facilities (and the belonging <BR>> objects) in the Praat software package". <P>Exactly. That is the advantage of having a program that models the <BR>physics (like Praat) instead of the mathematics (like Matlab). <P>Someone complained about the problem of choosing the "epsilon" in <BR>a Matlab implementation of the cepstrum. In a Praat script, you <BR>can express this epsilon in physical units. So if you want to add <BR>white noise with a density of 10^-8 Pascal/Hertz, your cepstrum <BR>routine would be (...)</BLOCKQUOTE> hi list, <P> just a brief comment about maths and physics... the advantage of having a program that deals directly with Physics units (as Praat's) is that the calibration is already done. If working with matlab, the mathematical models that render the physic behaviour must be proper calibrated to normalize the dynamic range (p.e: +32767/-32768 if your input is 16 bits quantifed) to a given value in a physic unit, say, for example, SPL units. If you feed a full-range 16 bit signal into a soundcard, output SPL depends ONLY on the gain applied to the amplifier that attacks the transducer (headph or loudspk). So, you can have a, let's say, 50 dB SPL signal with a +/- 32767 sine signal or with a +/- 16383 sine signal if the amplifier gain is modified properly. The difference between both cases is that the background noise increases but SPL pressure stands the same. <P> The problem of calibration is a very sensitive problem that is not always well managed and results coming from a incorrect calibration will probably give a wrong, even astonishing, result. <P> But please remember that physics modelling relies on a, let's say, 40%?empyrical data and a 60%? numerical modelling, so a math environment is well suited for physics models ONLY IF we use the correct *reality->virtuality* or *virtuality->reality* mapping. <P> Hope it helps and thank both to Praat and Matlab developers for providing us with such useful tools <P> have a nice day you all <P> Antonio <BR> <PRE>-- Antonio S. Pena E.T.S.E.Telecomunicacion Universidade de Vigo 36200 - VIGO ( Spain ) Phone: +34-986-812-130 FAX: +34-986-812-116 E-mail: apena(at)tsc.uvigo.es</PRE> </HTML> --------------28A7DFE5701F47FE2717839C--

This message came from the mail archive

http://www.auditory.org/postings/1999/

maintained by: DAn Ellis <dpwe@ee.columbia.edu>

Electrical Engineering Dept., Columbia University