Forums

temporal masking

Started by iulia March 16, 2010
Hello,

I want to implement in matlab the calculation of the masking threshold for an audio file. I read a lot about the frequential masking but I didn't find a lot of bibliography about temporal masking. It seems to be somethings simple but I realy don't understand how it works and how I integrate this in the global masking threshold.

Could you please help me with this... anything will be better than nothing!

Thanks,
Iulia
Iulia-

> I want to implement in matlab the calculation of the masking
> threshold for an audio file. I read a lot about the
> frequential masking but I didn't find a lot of bibliography
> about temporal masking. It seems to be somethings simple
> but I realy don't understand how it works and how I integrate
> this in the global masking threshold.
>
> Could you please help me with this... anything will be better
> than nothing!

I think it has to do with relative proximity of two sounds (how close together they occur). If you want to hide
(embed) signal A within signal B, then a likely place to do it is where signal B amplitude is high. Or if you want to
compress signal A, then more "compression effort" can be applied where signal A amplitude is low (because we can hear
details clearly), and less effort when signal A amplitude is high (wall of sound).

-Jeff
hi,

http://www.eas.asu.edu/~spanias/papers/paper-audio-tedspanias-00.pdf

this pdf explains a little about temporal masking,

mp3 standard ( iso 11172-3 ) calls it "pre echo control"

these should provide you with a good starting point.
Pradeep

.

On Tue, Mar 16, 2010 at 9:17 AM, Jeff Brower wrote:

> Iulia-
> > I want to implement in matlab the calculation of the masking
> > threshold for an audio file. I read a lot about the
> > frequential masking but I didn't find a lot of bibliography
> > about temporal masking. It seems to be somethings simple
> > but I realy don't understand how it works and how I integrate
> > this in the global masking threshold.
> >
> > Could you please help me with this... anything will be better
> > than nothing!
>
> I think it has to do with relative proximity of two sounds (how close
> together they occur). If you want to hide
> (embed) signal A within signal B, then a likely place to do it is where
> signal B amplitude is high. Or if you want to
> compress signal A, then more "compression effort" can be applied where
> signal A amplitude is low (because we can hear
> details clearly), and less effort when signal A amplitude is high (wall of
> sound).
>
> -Jeff
>
>
>