[vorbis] Low sample rates / bit rates

Shawn Riley roleypup at yahoo.com
Tue Nov 21 20:02:30 PST 2000



    I wrote:
> >Hey guys. I think Vorbis is pretty cool, but since the current OggEnc only offers 44.1kHz,
  Michael Smith wrote:
> Wrong. Oggenc supports different sample rates correctly. 
  Oops. My stupidity for not actually trying it. *embarrased blush* Even so, it's a little
misleading to have "OggEnc input files must currently be 44.1kHz, 16 bit stereo WAV files." on the
help screen. Perhaps it should read something like "Best quality will be achieved with 44.1kHz, 16
bit stereo WAV files". I just tried encoding something with the latest OggEnc, & it's a fair bit
faster than with Lame. The only advantage it seems that Lame has now, is that it supports mono
encoding & OggEnc does not.

> Libvorbis itself also supports different sample rates,  but is tuned for 44.1kHz, so other
things aren't handled optimally yet (the information on sample rates is fed through to the encode
engine, it just doesn't do all that much with it yet).
  I guess that's what I meant.

> Last I checked, lame still didn't have updated vorbis support - has this changed recently? If
not, it's possibly worse than oggenc at different sample rates. 
  I don't think it has. It most probably is worse. The compile that I was using was a prehistoric
one from - http://doslame.nm.ru/ - run by someone calling themself "tofikman". But in optimising
OggEnc specifically for 44.1k, is it likely to reduce quality for other sample rates compared to
before optimisation?

> finishing off the 1.0 feature list is higher in priority than tuning for other sampling
frequencies, I think. 
  Okay, you guys know what you're doing. Optimising for different sample rates sounds like a major
overhaul. I noticed that at lower bitrates, MP3 gets a lot of obvious artifacts that make it sound
messy, whereas Vorbis tends to sound more dry, & a lot more clean, which I think is a good thing.
Is that what's supposed to happen, or are obvious artifacts there that I can't hear? BTW, where's
the 1.0 feature list?

> What does -v4 do? I thought lame just ignored all the options when encoding vorbis. 
  I actually meant "-V4". Here's what Lame does with the "-V?" switch (which may have been changed
recently) -
* "lame --ogg -V0" => Vorbis mode E
* "lame --ogg -V1" => Vorbis mode C
* "lame --ogg -V4" => Vorbis mode B
* "lame --ogg -V5" => Vorbis mode A
* V2, V3 - ?... V6-V9 - mode A
  I think "--ogg" assumes "-v", which in turn assumes "-V4", so Lame probably runs Vorbis mode B
by default. And unfortunately, there's no way to call mode D with the "-V?" switch. From memory
(this isn't my home computer) Lame takes note of the "--resample", "-mm", "-a", & "-b" switches,
as well as some id3 switches, although, a compile of OggLame that I used previous to that (which
had no MP3 output) ignored id3 switches. Oh, what does Vorbis call what MP3 calls "id3"?

Shawn
roleypup at yahoo.com

__________________________________________________
Do You Yahoo!?
Yahoo! Shopping - Thousands of Stores. Millions of Products.
http://shopping.yahoo.com/

--- >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.



More information about the Vorbis mailing list