I know speex is not supposed to do a great job compressing music, but
I've noticed that the new VBR code chokes completely when you try to
compress horns. I've placed a particularly offensive example up at
http://www.utdallas.edu/~matthias/ . Take a look at a-16m*{ogg,spx}.
a.ogg is the first minute of an ogg created from the source media (in
44khz stereo). The rest have been mixed down to mono and resampled to
16khz. The fixed-bitrate recording at quality 5 is fairly listenable,
but the VBR encoding drops the bitrate down to where the artifacts
make it almost entirely unlistenable.
--
Matthias Granberry
matthias@utdallas.edu
(469) 371-0596
--- >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
'speex-dev-request@xiph.org'
containing only the word 'unsubscribe' in the body. No subject is
needed.
Unsubscribe messages sent to the list will be ignored/filtered.
On Sat, Jan 25, 2003 at 06:14:09AM -0600, Matthias Granberry wrote:> I know speex is not supposed to do a great job compressing music, but > I've noticed that the new VBR code chokes completely when you try to > compress horns. I've placed a particularly offensive example up at > http://www.utdallas.edu/~matthias/ . Take a look at a-16m*{ogg,spx}. > a.ogg is the first minute of an ogg created from the source media (in > 44khz stereo). The rest have been mixed down to mono and resampled to > 16khz. The fixed-bitrate recording at quality 5 is fairly listenable, > but the VBR encoding drops the bitrate down to where the artifacts > make it almost entirely unlistenable.Hi, my guess would be that in vbr mode, speex effectively finds all the bits that sound like speech, and compresses them really well, and then it throws the rest away, because it is a speech codec. without vbr it might be more generous with what it keeps :) Conrad. --- >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 'speex-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
The VBR code is really designed for speech in mind and any attempt to
make it better on music will likely make it worse on speech. If you
really want to encode music with Speex (why would you want to do
that??), you should not use VBR. Another thing to note: in lower
bit-rate wideband modes (which are sometimes turned on with VBR, but not
before), there's a processing that is completely unsuitable for music
and it might explain the problem.
Jean-Marc
Le sam 25/01/2003 à 07:14, Matthias Granberry a écrit :> I know speex is not supposed to do a great job compressing music, but
> I've noticed that the new VBR code chokes completely when you try to
> compress horns. I've placed a particularly offensive example up at
> http://www.utdallas.edu/~matthias/ . Take a look at a-16m*{ogg,spx}.
> a.ogg is the first minute of an ogg created from the source media (in
> 44khz stereo). The rest have been mixed down to mono and resampled to
> 16khz. The fixed-bitrate recording at quality 5 is fairly listenable,
> but the VBR encoding drops the bitrate down to where the artifacts
> make it almost entirely unlistenable.
--
Jean-Marc Valin, M.Sc.A.
LABORIUS (http://www.gel.usherb.ca/laborius)
Université de Sherbrooke, Québec, Canada
<p>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 242 bytes
Desc: signature.asc
Url :
http://lists.xiph.org/pipermail/speex-dev/attachments/20030127/2f8f4781/signature-0001.pgp