[vorbis] Vorbis native Ripper-decoder-encoder

[Gregory Maxwell]

On Sat, Aug 04, 2001 at 09:03:17PM +0200, Hongli Lai wrote:
> On 2001.08.04 20:23:25 +0200 Craig Dickson wrote:
> > ndrw mchl grnbrg wrote:
> > For someone with less-critical ears and storage limitations, if an MP3
> > file is too big and the original, non-encoded source is not available,
> > re-encoding it to a lower-bitrate MP3 is probably preferable to
> > transcoding to Vorbis, 
> 
> Why is that? How's MP3 better in this area?

MP3 is not better. However, since Vorbis has different a method for
windowing and blocking, quality will suffer much more then it would
otherwise.

To see an example of this, take a jpeg image, go through a cycle of
compressing it at 80% and then cropping one pixel off the top and left, wash
rinse repeat. Compare to the same number of cycles without the cropping.

There is always transcoding loss with perceptual formats, but there is much
more between different formats. In general, the more the formats differ, the
greater the transcoding loss.

If you have a 160kbit/sec mp3 which you would like to make 128kbit/sec, the
result would likely sound better with mp3 then with vorbis. If you have the
original, then the Vorbis result would be vastly superior.

Furthermore, much of the 160kbit/sec mp3 content out there bugs the hell out
of me, while I find 128kbit/sec vorbis to be acceptable. Were I to download
your resulting 128kbit/sec vorbis file, it's likely I'd be unpleasantly
surprised by the poor quality and would likely think less of Vorbis, because
of my ignorance of your bad practice.

--- >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.