Hi: I've been experimenting with streaming vorbis across my network from the windows box to an icecast2 server on my linux machine (the server I usually use is having some network issues at present). I've tested both ostream 0.7.1 and oddcast beta26. It seems that neither can encode on the fly on my P2-266 at 44100 fast enough to be able to stream it. Ostream looked the more promising of the two, with breaks in the audio but otherwise coping OK ... until it crashed, presumably because it filled some buffer somewhere. It can maintain a 44.1khz mono stream at lowest bitrate (aprox. 40kbps) but only just. I only tried oddcast once, at 64kbps stereo. It sounded very weird, as if it was only recording half the data in a given time period. It sounded as if it was playing at about double speed but wasn't pitch shifted. What I want to know is this. Am I expecting too much of my P2-266 to do 44.1khz 64kbps stereo in real-time? Assuming I'm not, is this an optimisation issue or should I be able to do it already? BTW: I don't know if anyone's fiddled with it or not, but I think results at lower sampling rates could be improved if the lowpass wasn't so severe. Of course, I don't know if this will introduce artifacts, but I can easily hear the roll-off and it's noticeably lower than the uncompressed audio would be at that rate. For example, the difference between vorbis at 44.1khz and 32khz is more noticeable, particularly in stereo, than it is for uncompressed samples. Geoff. --- >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@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, 8 Sep 2001, Geoff Shang wrote:> What I want to know is this. Am I expecting too much of my P2-266 to do > 44.1khz 64kbps stereo in real-time? Assuming I'm not, is this an > optimisation issue or should I be able to do it already?The best I've ever been able to do on my P2-266 at any bitrate is about 0.7RT using libvorbis 1.0RC2 and oggenc. I believe you need a P2-400 at least to do real-time encoding. -- Ignacio Vazquez-Abrams <ignacio@openservices.net> --- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Dude... you are going to need more CPU power, hands down. I have a stream up that is running off a 500Mhz Celeron with 256MB of ram and it can barely keep things going at 96kbps. I am using oddcast DSP Beta 26, and I do not have any problems with what you are describing below with the sound. In a very old version of oddcast there was a problem with similar to this. Check to make sure you have the most recent version , and try it again. For an example of a set up using oddcast, winamp, and icecast2win32 (also from oddsock) hit http://www.djlithium.com:8064/djlithium.ogg. Cheers. Lithium ----- Original Message ----- From: "Geoff Shang" <gshang@uq.net.au> To: <vorbis@xiph.org> Sent: Friday, September 07, 2001 7:03 AM Subject: [vorbis] Realtime encoding at 44.1khz> Hi: > > I've been experimenting with streaming vorbis across my network from the > windows box to an icecast2 server on my linux machine (the server Iusually> use is having some network issues at present). I've tested both ostream > 0.7.1 and oddcast beta26. It seems that neither can encode on the fly on > my P2-266 at 44100 fast enough to be able to stream it. Ostream lookedthe> more promising of the two, with breaks in the audio but otherwise copingOK> ... until it crashed, presumably because it filled some buffer somewhere. > It can maintain a 44.1khz mono stream at lowest bitrate (aprox. 40kbps)but> only just. I only tried oddcast once, at 64kbps stereo. It sounded very > weird, as if it was only recording half the data in a given time period. > It sounded as if it was playing at about double speed but wasn't pitch > shifted. > > What I want to know is this. Am I expecting too much of my P2-266 to do > 44.1khz 64kbps stereo in real-time? Assuming I'm not, is this an > optimisation issue or should I be able to do it already? > > BTW: I don't know if anyone's fiddled with it or not, but I think results > at lower sampling rates could be improved if the lowpass wasn't so severe. > Of course, I don't know if this will introduce artifacts, but I can easily > hear the roll-off and it's noticeably lower than the uncompressed audio > would be at that rate. For example, the difference between vorbis at > 44.1khz and 32khz is more noticeable, particularly in stereo, than it is > for uncompressed samples. > > Geoff. > > > > --- >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@xiph.org' > containing only the word 'unsubscribe' in the body. No subject is needed. > Unsubscribe messages sent to the list will be ignored/filtered. >--- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
At 12:03 AM 9/8/01 +1000, you wrote:>Hi: > >I've been experimenting with streaming vorbis across my network from the >windows box to an icecast2 server on my linux machine (the server I usually >use is having some network issues at present). I've tested both ostream >0.7.1 and oddcast beta26. It seems that neither can encode on the fly on >my P2-266 at 44100 fast enough to be able to stream it. Ostream looked the >more promising of the two, with breaks in the audio but otherwise coping OK >... until it crashed, presumably because it filled some buffer somewhere. >It can maintain a 44.1khz mono stream at lowest bitrate (aprox. 40kbps) but >only just. I only tried oddcast once, at 64kbps stereo. It sounded very >weird, as if it was only recording half the data in a given time period. >It sounded as if it was playing at about double speed but wasn't pitch >shifted. > >What I want to know is this. Am I expecting too much of my P2-266 to do >44.1khz 64kbps stereo in real-time? Assuming I'm not, is this an >optimisation issue or should I be able to do it already? >With the current code - yes. Based on my experiences (i.e. developing ices2), you'll need about a p2-333 at an absolute minimum to be able to do realtime encoding with the current libvorbis - it's certainly optimisable, though. However, if you're having problems with realtime encoding of a mono stream on a p2-266, then something else is wrong. For reference, ices2 takes about 60-70% of the cpu for encoding a stereo stream at 44.1kHz on my celeron 450. Mono is about half this - so your hardware shouldn't have any problems. Make sure you haven't accidently ended up with a debug build of libvorbis, instead of an optimised one - that'll kill performance. If you really can't even manage mono reliably and with plenty of spare cpu, then something else is wrong with the program you're using. Michael --- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.