[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: About importance of "phase" in sound recognition



Gain, I think you're missing the point. You could just as well use sine functions, I'm only using the FFT as a fast way to do a bunch of sines.

I'm not sure how you can argue it's the wrong test, when in fact you can show quite nicely that the modulation is audible.  I think if you tried the test, you'd find the sensoria goes beyond modulation, though.

I have no doubt that using different phases in low-frequency harmonic series will also provide a detectable output.  There really is no doubt about the idea of onset detection at low frequencies, is there?

__________________________
James D. Johnston  (jj@xxxxxxx)
CHIEF SCIENTIST - DTS, Inc.

425-522-0632 - office
425-814-3204 - fax
206-321-7449- mobile

11410 NE 122nd Way,  Suite 100
Kirkland, WA 98034

This electronic transmission (and/or the documents accompanying it) may contain confidential and privileged information. Any unauthorized use, copying or distribution is prohibited.  If you have received this communication in error, please notify DTS, Inc immediately by telephone (425-814-3200) and destroy the original message. Messages sent to and from us may be monitored.


-----Original Message-----
From: AUDITORY - Research in Auditory Perception [mailto:AUDITORY@xxxxxxxxxxxxxxx] On Behalf Of Daniel Möhlmann
Sent: Monday, October 11, 2010 2:31 AM
To: AUDITORY@xxxxxxxxxxxxxxx
Subject: Re: [AUDITORY] About importance of "phase" in sound recognition

Dear list,

While I think almost all of what has been said in the past days by different people was correct, I found some of it very misleading.

James Johnston has proposed a test about phase, using 496 Hz, 500 Hz and 504 Hz. This seems to be the wrong test, because that way you are really ab-using similar sinusoid components to create amplitude modulation. These components cause a slow, very audible beating at about 4 Hz. This has little to do with timbre.

Also, doing an enormous  2^20 FFT transform (which is 23 minutes at 44.1 KHz!) of the two different signals to prove that they have the same amplitude spectra is quite an artificial test. Within that time, the ear will have performed thousands of frequency measurements, and will - of course - find the signal to be fluctuating.

I agree that generating long lasting sine-wave components and listening to them is the most straightforward way of demonstrating the effect of phase audibility, much more straightforward at this point than discussing the wiring of nerves in the cochlea. But this has to be tested with non-interfering harmonic components and their "relative phases", as pointed out by James Beauchamp. Something like 200 Hz, 400 Hz, 600 Hz, 800 Hz. As I understand it, this has been done many times now, confirming that phase is audible under certain circumstances.

Daniel

Notice:
This message and any included attachments are intended only for the use of the addressee, and may contain information that is privileged or confidential. If you are not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please destroy the original message and any copies or printouts hereof.