[Vorbis-dev] Getting masked FFT data out of libvorbisenc

[Steinar H. Gunderson]

On Mon, Apr 23, 2007 at 05:41:40PM -0400, xiphmont at xiph.org wrote:
> If you want pure masking data, modify the code to not apply the ATH
> curve.  The only warning there is that the highest frequency masking
> data is somewhat more approximate than the rest as the ATH dominates
> and you'd never really get to see true HF masking except at very high
> SPL.

Thanks for the answer -- it looks like it could be an interesting metric (at
least research-wise) after that the modification. Correct me if I'm wrong
here, though: Can I expect this curve to be reasonable similar (in the sense
that the original signal is also similar) for the PCM original and something
that has already been through lossy compression once? (Just using the masking
curve, without the ATH curve removed, seems to give better values than pure
MFCC for most cases _except_ the case where it's comparing a FLAC original
and an MP3/Vorbis compressed version, which seems a bit odd to me.)

/* Steinar */
-- 
Homepage: http://www.sesse.net/