[vorbis] RC2 worse than RC1 and Beta4

[Monty]

On Thu, Aug 16, 2001 at 11:18:21PM -0400, M.r. Rage wrote:
>  >> Both
>  >> encoders SEEM to have a couple of dB bump at 10kHz.
>  >Not quite. You'll see a slight bump in rc2 at about 11kHz due to
>  >unnormalized point stereo. b4 did not perform coupling.
> 
> I did a little test to try to figure out what I'm hearing. Took a wav file 
> I made with 8kHz, 10kHz and 12kHz sine wave test tones (all the same 
> amplitude) and converted it to ogg 128k b4. Then I flipped the phase of the 
> ogg and added it to the original wav, figuring this would show the 
> artifacts induced in the compression. Looking at the resulting waveform, 
> 10k was about twice the average peak amplitude of the 8k, with 12k about 
> halfway in between. 

That methodology is flawed.  You will see an increase in noise in a
ramp at about 11kHz, following the ATH, because the ear is less
sensitive there.  This does *not* imply a net increase in
spectral energy.

(A trivial counterexample to your claim: If the signal is at +0dB, and
the noise signal you construct shows an increase in noise from -50dB
to -45dB at 11kHz, that does *not* indicate an increase of 5dB at
11kHz.  It indicates an increase of a less than a hundredth of a dB.
The log scale does not add linearly.  In addition, that only counts if
the noise is uncorrelated with the original signal, which, in this
case, it is not.  A net *decrease* in original signal strength at
11kHz would appear in your test as an 'increase' because it would
appear as out of phase 'noise'.)

Monty

--- >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.