My branch that I've been working on for the last few weeks is now the mainline; once I add a few more modes (mode_A.h through mode_E.h, corresponding roughly to 128,160,192,256,350 kbps stereo) and Mike Smith and Chad get the modes into OggEnc, we'll be at beta 2. A few things to note: I'd been really really hoping that I'd squeeze the last bit of performance out of non mid/side stereo to handle correlated noise (eg, voice) as smoothly as LAME/mp3 with mid/side. It's just not going to happen. Compare them in mono and Vorbis wins hands down. With m/s, however, the resolution boost to mp3 is too much to catch. Of course, in a month or so, we'll have mid/side stereo and then the winner will be clearly Vorbis across the board. I'd simply hoped to win on non-coupled stereo alone. Oh, well. (OTOH, the fact that a 64kbps mono Vorbis stream sounds as good as a 80kbps mono mp3 is cause for happiness. I just wanted more :-) Also, the current modes setup is a temporary hack. For one thing, submitting a mode description in a huge struct hardwired into the lib will present a binary compatability straightjacket. I'll be changing that aspect of the API. Lastly, I won't have all the modes I wanted in by noon tomorrow (eg, 96kbps isn't there yet). Hopefully all the others will be. Monty --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/
--- Monty <xiphmont@xiph.org> wrote:> > My branch that I've been working on for the last few > weeks is now the > mainline; once I add a few more modes (mode_A.h > through mode_E.h, > corresponding roughly to 128,160,192,256,350 kbps > stereo) and Mike Smith and > Chad get the modes into OggEnc, we'll be at beta 2. > > A few things to note: > > I'd been really really hoping that I'd squeeze the > last bit of performance out > of non mid/side stereo to handle correlated noise > (eg, voice) as smoothly as > LAME/mp3 with mid/side. It's just not going to > happen. Compare them in mono > and Vorbis wins hands down. With m/s, however, the > resolution boost to mp3 is > too much to catch. Of course, in a month or so, > we'll have mid/side stereo > and then the winner will be clearly Vorbis across > the board. I'd simply hoped > to win on non-coupled stereo alone. Oh, well. > > (OTOH, the fact that a 64kbps mono Vorbis stream > sounds as good as a 80kbps > mono mp3 is cause for happiness. I just wanted more > :-) > > Also, the current modes setup is a temporary hack. > For one thing, submitting > a mode description in a huge struct hardwired into > the lib will present a > binary compatability straightjacket. I'll be > changing that aspect of the API. > > Lastly, I won't have all the modes I wanted in by > noon tomorrow (eg, 96kbps > isn't there yet). Hopefully all the others will be. > > Monty > > > > --- >8 ---- > List archives: http://www.xiph.org/archives/ > Ogg project homepage: http://www.xiph.org/ogg/__________________________________________________ Do You Yahoo!? Yahoo! Mail – Free email you can access from anywhere! http://mail.yahoo.com/ --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/