>
> Hi,
>
> Can someone tell me where the minval values in the psychoacoustic model are
> based on? Or are they just vaues found by trial and error?
> Thanks
>
> Menno
>
They are handed down from the ISO in a table with no explanation. But
check:
Brandenburg and Johnston, "Second Generation Perceptual Audio
Coding: the hybrid code", AES 89th convention 1990.
I believe this paper outlines the tonality measure used in MP3.
The amount of masking is a function of the tonality, and I think the
minval's give a minimum masking for very atonal signals.
Mark
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
