Gregory Maxwell
2003-Jun-10 13:30 UTC
[vorbis-dev] Calling for 5.1 Mastering experience! (vorbis ambisonics and 5.1)
I've been doing a fair amount of work with Vorbis support for Ambisonics, which seems to be going along nicely. It seems that there is signifant interest in coding 5.1 material with Vorbis esp as tarkin becomes more complete, so I've decided to take a break from pure ambisonic work to look into this. I've decided that the best (from a pure elegance and patent avoidance) way to handle this is to basically decode the 5.1 input into a WXY or WXYUV ambisonic signal, which can then be handled by the Vorbis Ambisonic support. To accomplish that I produced a model of an idealized 5.1 speaker arrangement (with pointsource massless speakers) and a simulated ambisonic microphone at the listening location. I then measured the ambisonic impulse response from each speaker and produced convolution filters for the audio. The nice thing about this is that you can take the ambisonic decoded output from vorbis and convolve it with inverse of these filters and get pretty much the orignal input back or you could also convolve it to match your own speaker layout and get better results than you would with the orignal 5.1 material. My only problem right now is with the LFE (subwoofer channel). I had been intending to just mix it into the W signal, which I think matches the intent of the channel. Obviously this really couldn't be recovered on decode. My alternative is to use a seperate LFE channel in the vorbis stream, which is just ugly. So, what I'm looking for is some guidance about how the LFE is normally and best used in 5.1 mastering. If I had an infinite supply of perfectly flat really loud full spectrum speakers and I was setting up an ideal 5.1 setup, would I mix my LFE into my five speakers, or would I grap another one of those perfect speakers and lay it in the corner with the LFE feed going into it? <p><p>--- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Ray Heasman
2003-Jun-10 15:10 UTC
[vorbis-dev] Calling for 5.1 Mastering experience! (vorbis ambisonics and 5.1)
Hi, I am not an expert in this, but my sound engineer friend suggests this link as a potential answer to your question: http://www.mkprofessional.com/bass_mgmt.html Hope this helps, Ray On Tue, 2003-06-10 at 13:30, Gregory Maxwell wrote:> I've been doing a fair amount of work with Vorbis support for Ambisonics, > which seems to be going along nicely. It seems that there is signifant > interest in coding 5.1 material with Vorbis esp as tarkin becomes more > complete, so I've decided to take a break from pure ambisonic work to look > into this. > > I've decided that the best (from a pure elegance and patent avoidance) way > to handle this is to basically decode the 5.1 input into a WXY or WXYUV > ambisonic signal, which can then be handled by the Vorbis Ambisonic support. > > To accomplish that I produced a model of an idealized 5.1 speaker > arrangement (with pointsource massless speakers) and a simulated ambisonic > microphone at the listening location. I then measured the ambisonic > impulse response from each speaker and produced convolution filters for > the audio. > > The nice thing about this is that you can take the ambisonic decoded > output from vorbis and convolve it with inverse of these filters and get > pretty much the orignal input back or you could also convolve it to match > your own speaker layout and get better results than you would with the > orignal 5.1 material. > > My only problem right now is with the LFE (subwoofer channel). I had been > intending to just mix it into the W signal, which I think matches the > intent of the channel. Obviously this really couldn't be recovered on > decode. > > My alternative is to use a seperate LFE channel in the vorbis stream, > which is just ugly. > > So, what I'm looking for is some guidance about how the LFE is normally > and best used in 5.1 mastering. If I had an infinite supply of perfectly > flat really loud full spectrum speakers and I was setting up an ideal 5.1 > setup, would I mix my LFE into my five speakers, or would I grap another > one of those perfect speakers and lay it in the corner with the LFE feed > going into it? > > > > --- >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@xiph.org' > containing only the word 'unsubscribe' in the body. No subject is needed. > Unsubscribe messages sent to the list will be ignored/filtered.-- Ray Heasman <nurf@spamcop.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-dev-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.
Ralph Giles
2003-Jun-10 15:55 UTC
[vorbis-dev] Calling for 5.1 Mastering experience! (vorbis ambisonics and 5.1)
On Tuesday, June 10, 2003, at 09:30 pm, Gregory Maxwell wrote:> I've been doing a fair amount of work with Vorbis support for > Ambisonics, > which seems to be going along nicely. It seems that there is signifant > interest in coding 5.1 material with Vorbis esp as tarkin becomes more > complete, so I've decided to take a break from pure ambisonic work to > look > into this.I assume you're aware of the technical documentation on dolby's site? (http://www.dolby.com/pro/) In particular the surround mixing guide has a lot of detailed guidelines. I don't have any practical experience with it though, so I can't vouch for it. In particular it says that the LFE track is for option reinforcement, so while the idea is to encode it separately and mix it into the subwoofer channel, mixing it into the other 5 is acceptable, as is ignoring it. Don't know how safe that is in practice. Encoding it separately breaks the symmetry of the ambisonic encoding, but is probably closer to the intent of the original mix. It should compress quite well with it's own codebook given the lowpass, but I think it also tends to share a lot of entropy with the W channel.> I've decided that the best (from a pure elegance and patent avoidance) > way > to handle this is to basically decode the 5.1 input into a WXY or WXYUV > ambisonic signal, which can then be handled by the Vorbis Ambisonic > support.So U and V are the planar quadupole (m=+/-2) moments? I'm glad to hear you can get by with the same number of channels. Cheers, -r --- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Dave Mercier
2003-Jun-10 22:52 UTC
[vorbis-dev] Calling for 5.1 Mastering experience! (vorbis ambisonics and 5.1)
>So, what I'm looking for is some guidance about how the LFE is normally >and best used in 5.1 mastering. If I had an infinite supply of perfectly >flat really loud full spectrum speakers and I was setting up an ideal 5.1 >setup, would I mix my LFE into my five speakers, or would I grap another >one of those perfect speakers and lay it in the corner with the LFE feed >going into it?I can't advise too much from a sound design stand point, but I do have a fair amount of experience with this on the technology side.>From what I can tell, the LFE channel is a completely retarded invention. Inspeaking with members of the industry who do 5.1 as a business, they seem to think it's retarded too. In your case you would be better off with 5 full spectrum speakers. I think the only possible disadvantage would be that low bass coming from 5 speakers where you are not sitting in the sweet spot might have them all arriving out of phase and muddying your perception. Of course with all the bass coming out of the LFE, it can then be out of phase with the other 5 speakers. But if it's just high frequency content coming out of the other 5 speakers, it probably won't get as muddy as the low bass would. The following things make the LFE channel stupid: 1) In a standard fold-down, If the 5.1 signal is down mixed to anything lower like PLII, stereo, etc., the signal in the LFE channel is completely thrown away. So you must not put anything in the signal that is important. You can only put stuff in there that's not important. So I start to ask what's the point? 2) All receivers perform bass management. If you have pathetic small speakers and a kicking subwoofer, your receiver will implement a cross-over for you and filter all the bass content out of all channels and send it just to the subwoofer. If you have full range speakers it may not send anything to the sub woofer and leave the bass content in the mains. So the .1 channel is basically emulated half the time anyway. How this all fits into Ambisonics I no know. You might have to follow along and have a .1 channel anyway. I know the .1 channels makes other things hard as well, I wish the industry would just dump it (lets go for 8.0 or 10.0 next, not 7.1 or 8.2!). _________________________________________________________________ Add photos to your messages with MSN 8. Get 2 months FREE*. http://join.msn.com/?page=features/featuredemail --- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
ChristianHJW
2003-Jun-11 14:45 UTC
[vorbis-dev] Re: Calling for 5.1 Mastering experience! (vorbis ambisonics and 5.1)
Gregory Maxwell wrote:> I've been doing a fair amount of work with Vorbis support for Ambisonics, > which seems to be going along nicely. It seems that there is signifant > interest in coding 5.1 material with Vorbis esp as tarkin becomes more > complete, so I've decided to take a break from pure ambisonic work to look > into this.Great to see there is work done on Vorbis 5.1 !!! Latest AAC encoders were slowly taking Vorbis position in the DVD backup scene, because AAC 5.1 sounds quite nice at 200 - 240 kbps. About your question : From an sound engineering perspective, a separate subwoofer channel is more or less useless, it could be muxed fine into the main Left and Right speakers, resulting in a 5.0 setup. Never mux the signal into all channels, especially not the rear channels, as those speakers are normally - not built to be able to generate low frequency stuff at all - will act against the output of the main speakers, unless you inverse the phase by 180° , at least in a finite space = room. If you do inverse, to sum up the bass output from rear and front speakers, be sure there are a couple of setups where the 'rear' speakers are placed somewhere in the room where they found space, but not really in the rear, giving you problems again ;-) ... About 5.1 vs. 5.0 again : Only in very small rooms, where low frequencies can not even match the lambda/2 criterium from one wall to another ( again : finite space = room problem ) it could be possible that a subwoofer, placed in one corner of the room, could result in better low frequency performance. Just my 2 cents :) Christian <p>--- >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@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Christian HJ Wiesner
2003-Oct-24 06:45 UTC
[vorbis] Re: Calling for 5.1 Mastering experience! (vorbis ambisonics and 5.1)
Hi all, Hi Gregory, is there any progress on the surround encoder in Vorbis been done recently ? Christian <p>ChristianHJW wrote:> Gregory Maxwell wrote: > >> I've been doing a fair amount of work with Vorbis support for >> Ambisonics, >> which seems to be going along nicely. It seems that there is signifant >> interest in coding 5.1 material with Vorbis esp as tarkin becomes more >> complete, so I've decided to take a break from pure ambisonic work to >> look >> into this. > > > Great to see there is work done on Vorbis 5.1 !!! Latest AAC encoders > were slowly taking Vorbis position in the DVD backup scene, because > AAC 5.1 sounds quite nice at 200 - 240 kbps. > > About your question : > > From an sound engineering perspective, a separate subwoofer channel is > more or less useless, it could be muxed fine into the main Left and > Right speakers, resulting in a 5.0 setup. > > Never mux the signal into all channels, especially not the rear > channels, as those speakers are normally > > - not built to be able to generate low frequency stuff at all > - will act against the output of the main speakers, unless you inverse > the phase by 180° , at least in a finite space = room. If you do > inverse, to sum up the bass output from rear and front speakers, be > sure there are a couple of setups where the 'rear' speakers are placed > somewhere in the room where they found space, but not really in the > rear, giving you problems again ;-) ... > > About 5.1 vs. 5.0 again : > > Only in very small rooms, where low frequencies can not even match the > lambda/2 criterium from one wall to another ( again : finite space = > room problem ) it could be possible that a subwoofer, placed in one > corner of the room, could result in better low frequency performance. > > Just my 2 cents :) > > Christian ><p>--- >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.