Auditory JND

[jennifer resnik]

I am interested in calculating the auditory JND for different frequencies.
Does anyone have a matlab script to test auditory JND that I can use to start?

Thanks ,

Jennifer

[Wayne King]

"jennifer resnik" <jennife...@weizmann.ac.il> wrote in message <hhhtsk$22m$1...@fred.mathworks.com>...

Hi Jennifer, I don't have a handy script for frequency JNDs, but I know there are several publicly available Matlab toolboxes for psychoacoustic experiments.

http://www.psy.unipd.it/~grassi/mlp.html

https://gilbert.med.kuleuven.be/web/index.php/Public:Software/APEX

It shouldn't be too difficult to program a frequency JND experiment. Are you just interested in some simple design, two-alternative forced choice for example?

Wayne

[jennifer resnik]

Thank you for your answer.

Yes, I was thinking of doing something like that.

Jennifer

[Wayne King]

"jennifer resnik" <jennife...@weizmann.ac.il> wrote in message <hhi9ho$cnh$1...@fred.mathworks.com>...

> Thank you for your answer.
>
> Yes, I was thinking of doing something like that.
>
> Jennifer

I think that should be pretty easy either using the tools others have written in Matlab, or building off of theirs. Here is another psychoacoustics toolbox in Matlab:

http://psychtoolbox.org/wikka.php?wakka=HomePage

Looks like several of these provide routines for estimating the psychometric function, which is probably the toughest part of programming the experiment. The link above claims to have a user group (I can't speak for our viable it is) where you may find a targeted audience of psychophysicists using Matlab. If you can't find help at any of those sites, and you want to email me directly and not burden the list with prolonging the thread, that's fine. I know some psychoacousticians from a former life that I may be able to put you in contact with for some basic paradigms.

Wayne

[ImageAnalyst]

Jennifer:
One thing I hope you're aware of before you start this is that the
volume (speaker response) is very dependent on frequency. Just take a
look at the frequency response of this high end ($4000) speaker
system:
http://www.soundandvisionmag.com/speakers/2950/infinity-classia-home-theater-speaker-system-test-bench-page3.html
You can see how the ability of the speaker to match volumes at two
close frequencies is not so great, and that will affect your
psychophysical measurements. You would need to measure this in order
to correct for it, but then you will also have to deal with the
frequency response of the instrument you're using to measure the
speaker output with (the microphone). So it can get very tricky if
you plan on doing an accurate job. How do you plan on handling this
issue?
Regards,
ImageAnalyst

[jennifer resnik]

thank you for the links, I will start with the toolbox and I'll see how it goes.

Thank you,

Jennifer

[jennifer resnik]

Thank you for your comments.
I have to check exactly what equipment I am going to use, but we have a lab in the department that does many experiments using sound so I will talk to them. But if you have any ideas I would be glad to hear them.

Thanks,

Jennifer

[ImageAnalyst]

On Dec 31, 12:02 pm, "jennifer resnik"

---------------------------------------------------
Jennifer:
You need to characterize/calibrate your equipment, both the speaker
and the microphone used to characterize the speaker. I don't know
details on how to do that but I'm sure your research will find the
methods. Then you basically need to boost or dampen the signal at the
two frequencies so that they are presented to the listener with the
same volume. This is easily done just by dividing or multiplying by a
factor once you know the two volumes of the uncorrected signals.
Regards,
ImageAnalyst

[jennifer resnik]

Thank you,

Jennifer