TO: Peter/Sonic Foundry:
RE: Forum topic "Channels in 5.1 bleed?"
I apologize in advance for the length of this post…
In the thread "Channels in 5.1 bleed?", you state Sonic Foundry followed convention regarding implementation of the surround panner in Vegas.
[“…Panning is between left and right and the surround field. Center is not part of this. This is not how the center channel is defined for use. The center is the "mono" channel of 5.1...I know, there are different opinions on how this "should" be done - even amongst the Surround experts. We followed convention on this. --Peter"]
My assertion was that, in fact, the Vegas implementation is almost opposite of the industry standard, and as such, significantly limited for professional use. It is definitely NOT simply for reinforcing the phantom L/R center. You seem to be confusing center channel USAGE (at the creative discretion of the eng/producers etc.) with proper surround panner implementation (the h/w or s/w method to provide full flexibility for the aforementioned creative placement choices). Perhaps you are also confusing Dolby Digital Surround (AKA 5.1) with Dolby Surround (AKA the analog phase matrixed Pro Logic), as your comments about the center being a mono reinforcement image of the L/R signals seem to indicate.
Once again: The accepted industry standard is three DISCRETE channels up front, with the option of adding L/R "divergence" to the center pan position. The amount of center channel appearing in the front L and R busses is controlled by the divergence setting, with higher divergence equaling more of the center-panned program material added to the L and R busses. This provides complete flexibility for the three main uses of the center channel (discrete center channel, center channel program is spread out to L/R, L/R is summed to center channel – or combinations thereof).
I suggested that you should take a look at any of SF's competition such as Pro Tools, Nuendo etc. or any surround-capable studio or live mixing console (Neve, SSL, API, ATI, Yamaha, Tascam…) to confirm this.
Instead, in your subsequent response, you insist once again that Vegas' method is the convention:
[“I am basing my comments on recommend standards I have read. If you have alternative reading on 5.1 panning models, I would be very interested in reading them. I have discussed this with the other engineers involved in this here. The choice we made were based on information available in print. If there is some defacto standard that is accepted, we are wide open to providing more surround pan models. – Peter”]
Per your request, at the end of this post I have pasted comments solicited from respected engineers and surround experts on this subject (Peter/SF--I am happy to provide the original emails and contact info, should you wish). I also provided links to two Dolby 5.1 Guidelines documents that have been readily available on the Dolby web site for many years.
[from earlier post “… I just cant see adding such a feature unless it was in widespread use. --Peter"].
Not only is it in widespread use, it has been the standard in surround mixing for well over a decade, as Jim Hilson of Dolby Labs points out below. It would be of benefit to SF, as well as to potential and current users of Vegas, if the actual industry standard (and, therefore, fully flexible) surround panner is implemented.
Vegas is an impressive program and hopefully, through open-minded reception to suggestions, errors and bugs reported on the forum, it will only get better.
Sincerely,
Steve Murphy
From: Jim Hilson, Senior Broadcast Audio Specialist, Dolby Laboratories (www.dolby.com):
“Please allow me to comment on the Vegas panning. The information from Sonic Foundry may be the way they do it, but it is definitely not the standard panning practice used today.
The panning style used for most surround productions on current audio consoles utilizes "film style" panning. This panning has been in use for at least 10 years and pans left to center to right. It is absolutely necessary for proper image placement in 5.1 channel mixes.
The panning method in use by Vegas appears to be conventional stereo panning as was common with Dolby Surround tracks. In this style of encoding the center channel was optional on the encode side. The decoder didn't know the difference between a signal that was panned to center on a left/right pan pot and one that was inserted into the center channel input. The decoded output was the same.
Does it matter? Definitely YES!
In the film style scenario the image will pan across the room regardless of the viewer position relative to the center channel. In the Vegas scnario, the image will be shifted to the same side of center as the listener's seating position. In this case, you end up with both a phantom center image for the material panned between left and right, and a hard center image for the material assigned to the center channel. The only way it sounds correct is if you are seated in the center. The whole purpose of using a center channel is to eliminate this problem of shifted phantom images in the sound field…
Regards,
Jim Hilson
Senior Broadcast Audio Specialist
Dolby Laboratories”
From Alan Silverman, Arf Digital, multiple Grammy-winning recording and mastering engineer (www.arfdigital.com):
“LCR divergence is a firmly entrenched standard in every film mix console on the planet. The 02R96/DM2000/DM-24 mixers all have it, Samplitude has it and so on. The typical divergence control works in conjunction with the LCR panner (or joy stick) and is continuously variable from 100 (all signal to C, none to L/R) down to 0 (no signal to C, phantom center via L/R) This would be the case when the LCR panner is set to center. The LCR distribution varied in accordance with the same divergence law as the panner is moved from left to right.
Best,
Alan Silverman”
From Glen O’Hara, Surround Mixer, FDS Labs and Post Production Professor at the Conservatory of Recording Arts:
“I would disagree with Peter's responses. Exactly WHO are the "Experts" they contacted in the surround mixing arena? EVERY platform I have used in surround mixing, including digital consoles, Protools, even analog desks like SSL J series, Neve 88R, Euphonix, etc. allow for total panning control between any and all of the three front channels - EXCEPT ONE: The Tascam TMD4000 only allows panning from L thru C to R, meaning NO phantom L/R center…
Mike Sokol, who has written on, and been recording & mixing in, surround for years, just visited our campus (Conservatory of Recording Arts - Tempe, AZ) with a new 5.1 mix from a live stage performance he recorded. In this mix he placed a "stereo" image of an on-stage trap kit between Center and Right, while placing the onstage conga player (with stereo miking) between Center and LeftI had to compliment Mike on this mix, being a very effective use of 5.1 for a live recording. Mike said he placed the 'overhead mic's' for the drum kit hard Center & hard Right, then panned the individual 'snare' 'tom' 'hi hat' etc. from Center to Right.
In many of my own surround mixes we used the "phantom" positions of L to C, and C to R very effectively. Also, on one recent mix we put the main lead vocal 80%R/20%C , then put a second, or counterpoint lead 80%L/20%C -- very effective. [And all above examples are impossible to do in Vegas—SM].
If Vegas does not support full control of L-C-R levels, it would be a HUGE oversight, and if I might otherwise be interested in their platform, could NOT use it for 5.1 mixing just for that reason.”
Glen O'Hara
Post Production Dep't., Conservatory of Recording Arts (www.cras.org)
Contributing Editor, Pro Audio Review (www.proaudioreview.com)
Surround mixing engineer, FDS Labs (www.fdslabs.com)”
And for those who are still wondering why all of this is important, the following are a few more examples of very common techniques in digital surround mixing. As implemented, the Vegas surround panner will not let you:
*Pan one mono track discretely across the front three channels (i.e. when fully over L, C or R speaker, it does not appear in any of the other respective channels) [absolutely typical in film fx/post production]
*Statically place a mono dialog track in the center speaker, with a much smaller percentage of its level also in the L and R fronts [typical in dialog for film]
*Create a sound panning field between the L and C speakers (with no bleed to the R) or between the C and R speakers (with no bleed to the L). [typical in surround music mixing and film post]
For further info, read http://www.dolby.com/tech/5.1_mixing.pdf and http://www.dolby.com/tech/L.mn.0002.5.1Guide.s.pdf . These two docs have been readily available on the Dolby web site for many years.
RE: Forum topic "Channels in 5.1 bleed?"
I apologize in advance for the length of this post…
In the thread "Channels in 5.1 bleed?", you state Sonic Foundry followed convention regarding implementation of the surround panner in Vegas.
[“…Panning is between left and right and the surround field. Center is not part of this. This is not how the center channel is defined for use. The center is the "mono" channel of 5.1...I know, there are different opinions on how this "should" be done - even amongst the Surround experts. We followed convention on this. --Peter"]
My assertion was that, in fact, the Vegas implementation is almost opposite of the industry standard, and as such, significantly limited for professional use. It is definitely NOT simply for reinforcing the phantom L/R center. You seem to be confusing center channel USAGE (at the creative discretion of the eng/producers etc.) with proper surround panner implementation (the h/w or s/w method to provide full flexibility for the aforementioned creative placement choices). Perhaps you are also confusing Dolby Digital Surround (AKA 5.1) with Dolby Surround (AKA the analog phase matrixed Pro Logic), as your comments about the center being a mono reinforcement image of the L/R signals seem to indicate.
Once again: The accepted industry standard is three DISCRETE channels up front, with the option of adding L/R "divergence" to the center pan position. The amount of center channel appearing in the front L and R busses is controlled by the divergence setting, with higher divergence equaling more of the center-panned program material added to the L and R busses. This provides complete flexibility for the three main uses of the center channel (discrete center channel, center channel program is spread out to L/R, L/R is summed to center channel – or combinations thereof).
I suggested that you should take a look at any of SF's competition such as Pro Tools, Nuendo etc. or any surround-capable studio or live mixing console (Neve, SSL, API, ATI, Yamaha, Tascam…) to confirm this.
Instead, in your subsequent response, you insist once again that Vegas' method is the convention:
[“I am basing my comments on recommend standards I have read. If you have alternative reading on 5.1 panning models, I would be very interested in reading them. I have discussed this with the other engineers involved in this here. The choice we made were based on information available in print. If there is some defacto standard that is accepted, we are wide open to providing more surround pan models. – Peter”]
Per your request, at the end of this post I have pasted comments solicited from respected engineers and surround experts on this subject (Peter/SF--I am happy to provide the original emails and contact info, should you wish). I also provided links to two Dolby 5.1 Guidelines documents that have been readily available on the Dolby web site for many years.
[from earlier post “… I just cant see adding such a feature unless it was in widespread use. --Peter"].
Not only is it in widespread use, it has been the standard in surround mixing for well over a decade, as Jim Hilson of Dolby Labs points out below. It would be of benefit to SF, as well as to potential and current users of Vegas, if the actual industry standard (and, therefore, fully flexible) surround panner is implemented.
Vegas is an impressive program and hopefully, through open-minded reception to suggestions, errors and bugs reported on the forum, it will only get better.
Sincerely,
Steve Murphy
From: Jim Hilson, Senior Broadcast Audio Specialist, Dolby Laboratories (www.dolby.com):
“Please allow me to comment on the Vegas panning. The information from Sonic Foundry may be the way they do it, but it is definitely not the standard panning practice used today.
The panning style used for most surround productions on current audio consoles utilizes "film style" panning. This panning has been in use for at least 10 years and pans left to center to right. It is absolutely necessary for proper image placement in 5.1 channel mixes.
The panning method in use by Vegas appears to be conventional stereo panning as was common with Dolby Surround tracks. In this style of encoding the center channel was optional on the encode side. The decoder didn't know the difference between a signal that was panned to center on a left/right pan pot and one that was inserted into the center channel input. The decoded output was the same.
Does it matter? Definitely YES!
In the film style scenario the image will pan across the room regardless of the viewer position relative to the center channel. In the Vegas scnario, the image will be shifted to the same side of center as the listener's seating position. In this case, you end up with both a phantom center image for the material panned between left and right, and a hard center image for the material assigned to the center channel. The only way it sounds correct is if you are seated in the center. The whole purpose of using a center channel is to eliminate this problem of shifted phantom images in the sound field…
Regards,
Jim Hilson
Senior Broadcast Audio Specialist
Dolby Laboratories”
From Alan Silverman, Arf Digital, multiple Grammy-winning recording and mastering engineer (www.arfdigital.com):
“LCR divergence is a firmly entrenched standard in every film mix console on the planet. The 02R96/DM2000/DM-24 mixers all have it, Samplitude has it and so on. The typical divergence control works in conjunction with the LCR panner (or joy stick) and is continuously variable from 100 (all signal to C, none to L/R) down to 0 (no signal to C, phantom center via L/R) This would be the case when the LCR panner is set to center. The LCR distribution varied in accordance with the same divergence law as the panner is moved from left to right.
Best,
Alan Silverman”
From Glen O’Hara, Surround Mixer, FDS Labs and Post Production Professor at the Conservatory of Recording Arts:
“I would disagree with Peter's responses. Exactly WHO are the "Experts" they contacted in the surround mixing arena? EVERY platform I have used in surround mixing, including digital consoles, Protools, even analog desks like SSL J series, Neve 88R, Euphonix, etc. allow for total panning control between any and all of the three front channels - EXCEPT ONE: The Tascam TMD4000 only allows panning from L thru C to R, meaning NO phantom L/R center…
Mike Sokol, who has written on, and been recording & mixing in, surround for years, just visited our campus (Conservatory of Recording Arts - Tempe, AZ) with a new 5.1 mix from a live stage performance he recorded. In this mix he placed a "stereo" image of an on-stage trap kit between Center and Right, while placing the onstage conga player (with stereo miking) between Center and LeftI had to compliment Mike on this mix, being a very effective use of 5.1 for a live recording. Mike said he placed the 'overhead mic's' for the drum kit hard Center & hard Right, then panned the individual 'snare' 'tom' 'hi hat' etc. from Center to Right.
In many of my own surround mixes we used the "phantom" positions of L to C, and C to R very effectively. Also, on one recent mix we put the main lead vocal 80%R/20%C , then put a second, or counterpoint lead 80%L/20%C -- very effective. [And all above examples are impossible to do in Vegas—SM].
If Vegas does not support full control of L-C-R levels, it would be a HUGE oversight, and if I might otherwise be interested in their platform, could NOT use it for 5.1 mixing just for that reason.”
Glen O'Hara
Post Production Dep't., Conservatory of Recording Arts (www.cras.org)
Contributing Editor, Pro Audio Review (www.proaudioreview.com)
Surround mixing engineer, FDS Labs (www.fdslabs.com)”
And for those who are still wondering why all of this is important, the following are a few more examples of very common techniques in digital surround mixing. As implemented, the Vegas surround panner will not let you:
*Pan one mono track discretely across the front three channels (i.e. when fully over L, C or R speaker, it does not appear in any of the other respective channels) [absolutely typical in film fx/post production]
*Statically place a mono dialog track in the center speaker, with a much smaller percentage of its level also in the L and R fronts [typical in dialog for film]
*Create a sound panning field between the L and C speakers (with no bleed to the R) or between the C and R speakers (with no bleed to the L). [typical in surround music mixing and film post]
For further info, read http://www.dolby.com/tech/5.1_mixing.pdf and http://www.dolby.com/tech/L.mn.0002.5.1Guide.s.pdf . These two docs have been readily available on the Dolby web site for many years.