[vorbis-dev] new vorbisenc behaviour

[Monty]

On Wed, Jan 02, 2002 at 01:26:55PM +0000, David Burnett wrote:

> I know about the new API but the old API is still there and I would still
> expect it to work as before, i.e. to quote the *RC3 oggenc code*

We intended to keep the behavior identical to old priginally, but it
would have caused anyone new to the scene to read the API and wonder
what we were smoking.  We changed it as little as possible to keep it
clean; I didn't like having deprecated setup functions before 1.0 even
went out the door, although this might have been a better solution.

> If the new version vorbis_encode_init(...) is not a finished article then
> fair enough, but if it is then approx_bitrate_to_vbr(...) is way out
> with its approximations.

It's a finished article for now.  The API will not change further.  As
for approximations, that's what they have to be: approximations.  If
they were correct, they'd be solutions, not approximations.  There are
better ways to do the bitrate bounding than I am doing; no question of
it.  But the design decision we made is that libvorbis is a VBR
encoding lib, and bitrate management is a secondary feature.  Perhaps,
come Vorbis II, that will change, but that's the way it is for now.

The bitrate management/tracking layer is quite good, but the way the
actual management is being effected, through a somewhat degenerate
peeling mechanism, is what is not optimal.  It is, however, very
simple.  An optimal mechanism would not be built on top of the VBR
coding.

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-dev-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.