peeling as I understand it (was Re: [vorbis] When will qualityincrease be unnoticable?)

[Alan MacDonald]

> -----Original Message-----
> From: owner-vorbis at xiph.org [mailto:owner-vorbis at xiph.org]On Behalf Of
> Ed Sweetman
>
> that's not peeling, that's just re-encoding
>
But the way the codebooks work, in a layered fashion of adding granularity
to the residue values at each pass, it actually works much better to
re-quantize than just peeling off a layer of the residue, I believe.  Your
msb to lsb analogy is basically what happens when encoding.  Lets say you
have a residue value of 156, there could be a codebook with values
between -100 and 100 with 5 steps  -100, -50, 0, 50, 100.  So that would be
the first pass, the second pass could have -50 to 50 with 11 steps and then
the final pass would be -5..5.  So lets say 156 quantizes to 155.  You could
peel off the -5..5 book and end up with 150 and then peel off the 2nd book
and end up with 100. My point is that it's not that much more work to decode
the codebooks and use new ones, your not doing any transforms or
psycho-acoustics, which is the most computationally intensive parts of the
encoding, your not really re-encoding, your just re-quantizing.

<p>Later,
Alan

<p><p><p>--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.