I've been asked to upmix some material for motion pictures in 3.0. Obviously, in a movie theatre, it would be weird to hear a classic recording with some musicians' parts coming from the rear.

A 3.0 upmix is identical to the original stereo, however the center channel (lead vocal, etc.) is eliminated from the front left/front right and channeled exclusively to a firm, real center. The overall mix is unchanged from the stereo, except that the center is real, and doesn't wander around the room as you do.

It also provides a format that is usable for those who don't have or want rear speakers.

Do you think you would ever prefer a 3.0 version of a mix, especially of legacy stereo mixes, that is, with everything up front?

=John

...........................especially of legacy stereo mixes, that is, with everything up front?no

Do you think you would ever prefer a 3.0 version of a mix, especially of legacy stereo mixes, that is, with everything up front?

no.

Hmm, interesting question.

I suppose, for purists, a question might be what happens in my receiver if I send it a 3.0 signal but tell it I have no center speaker and it mixes that center channel signal into the (front) left and right channels.

If a true 3.0 mix could replicate the original stereo (2-ch) mix in this manner, then theoretically all listeners could listen as they see fit.

The question is how well centered in the 2-ch mix are the vocals, etc that will get routed to the discrete center? There could be an argument made that the stuff that sounds like it is centered in a 2-ch mix isn't really perfectly centered, perhaps due to some studio trickery, or just using a stereo mike on purpose and allowing the singer to "drift".

I know you brought this up in the other thread about full-range center speakers, where the center channel can be equated by L and R signals 4.5 db lower (IIRC - feel free to correct me), and that mixing was commonly done "in the olden days" ;) using both what we regard today as 3.0 and 2.0 set-ups. But if today's mixes are developed in a 2.0-only environment, how truly anchored is that phantom center?

Really, I dunno. I wouldn't mind hearing it as an experiment, at the very least.

IMHO,

shinksma

This works okay in a situation where you have three matched speakers across the front and perfectly integrated with the room, but VERY few people actually have this. So I would guess most people would be against this idea for music.

Penteo is asking about 3.0 upmixes, specifically. Not native 3.0 mixes. I think many of us would view 3.0 upmixes and native 3.0 mixes very differently.

Conversion to 3.0 with a "hard" vocal center will of course limit your music choices considerably.
With a stereo to 5.0 or 5.1 conversion one is not quite as constrained, depending on what degree of "discreteness" you can be satisfied with.

A fair amount of '70s music, for example, used panning that carried the vocals (and instruments) from side to side- as I'm sure you well know.

But I guess for me, the short answer is "no". ;)

What I would love to see is Profile 3.0. ;)

How would this be different from the RCA Living Stereo 3.0 channel releases (which all sound great on my system, despite the fact that my centre channel speaker is NOT identical to the mains--they have the same tweeter and are of the same line)? If it is anything like the snippets I've heard of the 5.1 Penteo process, I think I would be willing to give it a try before dismissing the idea out of hand.

Obviously, in a movie theatre, it would be weird to hear a classic recording with some musicians' parts coming from the rear.But it wouldn't be weird to hear a classical recording where the recorded ambience comes from the surround field instead of up front. In fact, it would sound more natural. If you deliver a 3-channel mix, the ambient cues will still be in place in the recording, and movie mixers can use ambience extraction tools to pull as much or as little information into the surrounds as they want. To that end, delivering a 3-channel mix could sometimes be preferable to giving them a 5-channel mix. Do you think you would ever prefer a 3.0 version of a mix, especially of legacy stereo mixes, that is, with everything up front?If the centre content is extracted well, I would prefer it over 2-speaker playback every time. From the samples I've heard so far, Penteo does a better job of centre extraction than any of the real time processing I use.

With an extracted centre, content that was mixed to image at the middle of the soundstage no longer has to be reproduced as dual-mono, comb-filtering, phantom-imaged sounds. Plus, the soundstage becomes much more stable, with rock-solid centre imaging.

Sanjay

How would this be different from the RCA Living Stereo 3.0 channel releases...?Those were native 3-channel recordings, whereas John is talking about extracting centre content from 2-channel recordings.

Sanjay

Those were native 3-channel recordings, whereas John is talking about extracting centre content from 2-channel recordings.

Sanjay

I know. I should have been clearer. I meant "how is it different in terms of enjoying the benefits of the anchored centre information?" When I sit in the sweet spot and stay still, I can't really tell there's a major difference between the 2.0 and 3.0 versions of my RCA discs but when I cede "the chair" to my buddy who drops by every once in a while to spin a few classical discs, I find the experience on the couch (off to the side of "the chair") to be far more enjoyable in 3.0 than in 2.0 with those Living Stereo discs that offer a 3.0 mix. If 5.1 Penteo is anything to judge by, I suspect the benefits would be similar, even if the process by which the centre channel is generated differs.

Considering that I don't even use a center speaker for HT, much less music, I really can't see it.
Stereo is perfectly convincing on a good system, when seated in or near the sweet spot.

I know. I should have been clearer. I meant "how is it different in terms of enjoying the benefits of the anchored centre information?" When I sit in the sweet spot and stay still, I can't really tell there's a major difference between the 2.0 and 3.0 versions of my RCA discs..........................Really? That's one of my favorite demonstrations of the advantages of the discrete center channel and, chez moi, it has never failed to impress the subject, even those who are died-in-the-wool 2channel guys.

It may be convincing on music (as most music mixing is done with 2 channel in mind) but I have to disagree about the centre channel for movies. Nearly 60% of a 5.1 mix is in the centre and while 2 channel "downmixes" of 5.1 movie soundtracks are serviceable, they pale in comparison to having a centre channel. If, for some reason, I could not have rear speakers in my HT, at the very least I would configure my system as 3.1.

Really? That's one of my favorite demonstrations of the advantages of the discrete center channel and, chez moi, it has never failed to impress the subject, even those who are died-in-the-wool 2channel guys.

There are two things about my observation to keep in mind. One (and most important) is the keeping my head "perfectly still" part. Even minor shifts reveal the centre channel's advantage (it's why I always choose the 3.0 mix when it's available). The second one is that the improvement of the 3.0 vs 2.0 mix "from the sweet spot" is less dramatic, in relative terms, to the improvement when I am constrained to listen outside the "sweet spot". My overall point was, to me, the benefits of the 3.0 mix are even more apparent when I'm forced out of the "sweet spot", not that there was little or no benefit from the "sweet spot" itself.

It may be convincing on music (as most music mixing is done with 2 channel in mind) but I have to disagree about the centre channel for movies. Nearly 60% of a 5.1 mix is in the centre and while 2 channel "downmixes" of 5.1 movie soundtracks are serviceable, they pale in comparison to having a centre channel. If, for some reason, I could not have rear speakers in my HT, at the very least I would configure my system as 3.1.Are we disagreeing? I thought you were discounting the value of a discrete center when sitting on-axis. If not, good.

OTOH, the RCAs were recorded (down to) 3 channels at the sessions and then mixed to 2channels for the original releases. Same for the LS reissues and, in my experience, the 2channel mix-downs are simply not as good as the 3channel tracks. If you can A/B them, the difference is quite apparent.

Are we disagreeing? I thought you were discounting the value of a discrete center when sitting on-axis. If not, good.

OTOH, the RCAs were recorded (down to) 3 channels at the sessions and then mixed to 2channels for the original releases. Same for the LS reissues and, in my experience, the 2channel mix-downs are simply not as good as the 3channel tracks. If you can A/B them, the difference is quite apparent.

No, we're not disagreeing. That post was in response to PULLIAM who appears to believe the centre is not worth having for either movies or music. I was simply saying that as nice as it is with music, a centre channel is essential with a 5.1 movie soundtrack to get its full effect (in other words, while I prefer MCH audio for both movies and music, I am more prepared to sacrifice a centre channel for music, if necessary, though I would never suggest it as a preference).

As for the RCAs, I will do some more A/Bing (a good excuse to fire up the gear). When I made the observation above, I only tried it with 2 of the 30 or so that I own--perhaps the others will be more impressive in an A/B comparison.

Ultimately, though, the more MCH (hi-res, discrete and lossless wherever possible) music available, the better.

It may be convincing on music (as most music mixing is done with 2 channel in mind) but I have to disagree about the centre channel for movies. Nearly 60% of a 5.1 mix is in the centre...If you listen carefully to multi-channel music, you'll find that there is more content mixed to image at the centre of the front soundstage than anywhere else. This even holds true for 2-channel music: important stuff, like instrument solos and lead vocals, are almost always mixed to image dead centre. The difference is that movie mixes put centre imaged content in the centre channel whereas many multi-channel music mixes continue to split those sounds to the front L/R channels.

The 3-channel RCA recordings are much more like movie mixes: there is more in the centre channel than the L/R channels (which contain very little dual-mono content). That's one of the reasons Kal can demonstrate the difference between the 2.0 and 3.0 mixes right from the sweet spot. His centre speaker is moving lots of air at that location, which can't be duplicated by a phantom centre image that's created in the brain.

Sanjay

The 3-channel RCA recordings are much more like movie mixes: there is more in the centre channel than the L/R channels (which contain very little dual-mono content). That's one of the reasons Kal can demonstrate the difference between the 2.0 and 3.0 mixes right from the sweet spot. His centre speaker is moving lots of air at that location, which can't be duplicated by a phantom centre image that's created in the brain.
SanjayBingo!

BTW, one reason I keep an old and still functioning Philips SACD1000 (nicely tweaked by Alex Peychev of APL) is that it is, afaik, the only SACD player which can switch tracks (among CD/2channelSACD/multichannelSACD) ON-THE-FLY! Does anyone know of another?

That's one of my favorite demonstrations of the advantages of the discrete center channel.................The 3-channel RCA recordings are much more like movie mixes.............I disagree. The center information on these live recordings is NOT really that discrete.

These recordings were made simply with 3 mics ( I presume hanging) equidistant from the orchestra; left, center, and right. Other than setting the levels, there is no other manipulation. Exactly how the center mic's level was set relative to the others, I do not know. But when I listen to these discs in 3.0, the sound seems to come from the 3 front speakers in equal parts; the center channel is not dominant.

Sure, the microphones picked up different info while recording the orchestra, and each of the 3 channels IS different, but I would not consider them to be very discrete like that of a heavily manipulated movie soundtrack.

I do not think these recordings are that analogous to a modern movie soundtrack at all.

As an aside, these 3 channel recordings STILL do not approximate the sound of an orchestra THAT well and, honestly, I am not even entirely convinced that the 3.0 version is necessarily more "true" in that sense than the 2.0 downmix. Sounds better in your den or HT? Maybe. Approximates the sound of an orchestra better? I don't know.

I disagree. The center information on these live recordings is NOT really that discrete.Really, have you tried any of the Heifetz recordings in this series?

I do not think these recordings are that analogous to a modern movie soundtrack at all. There we agree.

As an aside, these 3 channel recordings STILL do not approximate the sound of an orchestra THAT well and, honestly, I am not even entirely convinced that the 3.0 version is necessarily more "true" in that sense than the 2.0 downmix. Sounds better in your den or HT? Maybe. Approximates the sound of an orchestra better? I don't know.I did not say that any of them approximate the sound of a real orchestra better better in 3.0 than in 2.0. Nonetheless, there are consistent advantages to the 3.0 including better delineation of the soloists and a greater, more natural spread of the orchestra. Harmonic and balance issues remain.

Really, have you tried any of the Heifetz recordings in this series?I have the Beethoven and Mendelssohn Violin Concerto disc. IIRC only one of those recordings is 3-channel. I'll give it a good listening to. I also have the Mercury Janis Rachmaninoff Piano Concertos which may demonstrate a more discrete center, too, although I'm not sure where the piano might have been located relative to the mics. If to either side, though, it may still be indiscreetly discrete (:D).


I did not say that any of them approximate the sound of a real orchestra better better in 3.0 than in 2.0.No one did. I said "as an aside............".

The center information on these live recordings is NOT really that discrete.

These recordings were made simply with 3 mics ( I presume hanging) equidistant from the orchestra; left, center, and right. Other than setting the levels, there is no other manipulation.In that case, is the left & right channel information in 2-channel recordings "NOT really that discrete"? when I listen to these discs in 3.0, the sound seems to come from the 3 front speakers in equal parts; the center channel is not dominantMy processor displays levels for every input channel, making it clear that the centre channel in these recordings contain more than the L&R channels.

However, the centre channel shouldn't 'sound' dominant. It certainly doesn't with movie soundtracks that place 60-70% of the information in the centre. In fact, the soundstage thrown by the 3-channel mix should have the same proportions and balance as the 2-channel version, which contains just as much centre imaged content (albeit placed in the L/R channels). I do not think these recordings are that analogous to a modern movie soundtrack at all.They're analogous to the extent that centre imaged content is loaded into the centre channel, not equally split equally into the L/R channels (as is typical of music mixes, 2-channel and multi-channel). these 3 channel recordings STILL do not approximate the sound of an orchestra THAT well and, honestly, I am not even entirely convinced that the 3.0 version is necessarily more "true" in that sense than the 2.0 downmixIt's not a question of how well 3-channels approximate an orchestra, the point it that it does a better job than 2 channels by overcoming several limitations (some of which have been mentioned in this thread). That's why "stereo" was originally 3 channels/speakers, not 2.

Sanjay

In that case, is the left & right channel information in 2-channel recordings "NOT really that discrete"?If they're live recordings with two mics with no other manipulation they are certainly less discrete than most manipulated studio recordings. I emboldened live recordings for a reason. But, yes, I have live recordings done this way that definitely cast a "stereo" soundfield.


It's not a question of how well 3-channels approximate an orchestra, the point it that it does a better job than 2 channels by overcoming several limitations (some of which have been mentioned in this thread). That's why "stereo" was originally 3 channels/speakers, not 2."Does a better job than 2 channel" at what? Approximating the sound of an orchestra? I'm not convinced. They sound different. It may be your opinion that it approximates the sound of the orchestra better. Although you may be able to show me the studies, I've not seen DBTs of 2 vs. 3 channel live recordings with "which approximates the sound of an orchestra better" as the criteria. This would be so highly subjective as well as so highly dependent on the recording methods that it would probably be impossible to do it correctly.

"Approximating the sound of an orchestra" is pretty much never one of my goals, or something I would use as a reference. Almost all of my CDs are studio recordings.

"Approximating the sound of an orchestra" is pretty much never one of my goals, or something I would use as a reference. Almost all of my CDs are studio recordings.You'll find some "studio recordings" of orchestras, but they're still (usually) recorded very similarly to the way they would be recorded live.

And obviously, yes, approximating the sound of an orchestra MAY not necessarily be THE goal, anyway, or even an achievable goal, for that matter.

But I think that for many live orchestral recordings (and even other live recordings), that IS an underlying goal.

"Does a better job than 2 channel" at what?Mimicking reality.

Every stereo recording has correlated mono content in the L/R channels that phantom images in the centre of the soundstage. In real life, this would be reproduced as mono sounds from that location, just as they are with a centre speaker. With 2-speaker playback, these sounds are reproduced as a dual-mono phantom images. You never hear it that way in real life, why reproduce it that way at home?

Centre imaged sounds in real life go straight to both your ears, just as they do with a centre speaker. With 2-speaker playback, those mono sounds are actually heard four times: the sound from each speaker is first heard in the nearby ear, with some frequencies then going around your head to the far ear. So centre imaged sounds are heard as two sets, separated by a short time interval. Those sounds are never heard that way in real life, why hear them that way at home?

BTW, this difference in how we hear a phantom centre vs a hard centre is why soundstage depth is reproduced more easily and heard more clearly with 3 channels than 2. The brain doesn't have to separate out the dual arrivals of centre image. And, as mentioned before, a centre speaker will pressurize the air at the location where those sounds are supposed to come from, just as would occur in real life. Phantom images are created in the brain and cannot reproduce that effect in the physical world.

All of the above is for a listener seated in the sweet spot. If you move off-axis, then no contest. When you have a vocal soloist standing in front of you, she'll occupy the same location in relation to the soundstage no matter where you're sitting. With 2 speaker playback, you move and the phantom centre image moves with you. You never have to deal with that sort of image drift in real life, why tolerate it at home? I'm not convinced.Doesn't matter. What's more important is whether you can refute the explanations that I and others have posted. All you've done so far is say you don't believe it.

Sanjay

Mimicking reality.What "reality"? For a live recording, there is a "reality" involved, but show me where it is proven that a 3-channel recording "mimick [sic] reality" in this instance better than a 2-channel recording? As for studio recordings, there really is no "reality", is there? Unless the recording is trying to "mimic", exactly, a live situation, in which case I can obviously see how a 3rd speaker can help a studio recording achieve this goal, but not necessarily because it "centers" anything.

But I do think it is important to distinguish between the differences between the goals of live versus studio recordings, here.


Every stereo recording has correlated mono content in the L/R channels that phantom images in the centre of the soundstage. In real life, this would be reproduced as mono sounds from that location, just as they are with a centre speaker. With 2-speaker playback, these sounds are reproduced as a dual-mono phantom images. You never hear it that way in real life, why reproduce it that way at home?

Centre imaged sounds in real life go straight to both your ears, just as they do with a centre speaker. With 2-speaker playback, those mono sounds are actually heard four times: the sound from each speaker is first heard in the nearby ear, with some frequencies then going around your head to the far ear. So centre imaged sounds are heard as two sets, separated by a short time interval. Those sounds are never heard that way in real life, why hear them that way at home?

BTW, this difference in how we hear a phantom centre vs a hard centre is why soundstage depth is reproduced more easily and heard more clearly with 3 channels than 2. The brain doesn't have to separate out the dual arrivals of centre image. And, as mentioned before, a centre speaker will pressurize the air at the location where those sounds are supposed to come from, just as would occur in real life. Phantom images are created in the brain and cannot reproduce that effect in the physical world.

All of the above is for a listener seated in the sweet spot. If you move off-axis, then no contest. When you have a vocal soloist standing in front of you, she'll occupy the same location in relation to the soundstage no matter where you're sitting. With 2 speaker playback, you move and the phantom centre image moves with you. You never have to deal with that sort of image drift in real life, why tolerate it at home?

What's more important is whether you can refute the explanations that I and others have posted.
It's all "technobabble" (even if it makes sense, it's still that). Where is there any proof at all that having a 3rd center speaker necessarily provides a "better", "truer", or more "real" listening experience?

All you've done so far is say you don't believe it.What is there to believe (or disbelieve)?



As expected, you've assumed an argumentative stance, here, sanjay. If you read my responses they are all quite objective and I have not taken any particular stand except to say that I would not like my stereo recordings to be "upmixed" to 3-channel (the topic of this thread, remember). I'm not arguing anything. I am leaving it wide open as to what may sound "best". But to make the blanket statement that 3-channel is always "better" is a stretch. Particularly since "better" is quite indefinable and nebulous in this case.

I would actually buy native 3-channel recordings, not because I think that the center channel necessarily provides "reality" or "centers" anything, but probably more because it doesn't. What would George Martin have done with the 3rd center channel in 1967? He wouldn't have used it to provide "reality" any more than Parson's uses quad mixing capability to provide "reality" in the DSoTM quad mix.

It's all "technobabble" (even if it makes sense, it's still that).Why don't you point out which parts of my explanation are not factual. That would certainly be more productive than attempting to dismiss it by calling it a name. I am leaving it wide open as to what may sound "best".We aren't discussing what "sounds best", because that is based on personal preference. I was simply answering your question by explaining what 3-channel recordings do better than 2-channel recordings. What would George Martin have done with the 3rd center channel in 1967?The answer is in the recent 'Love' DVD-A, which he and his son Giles mixed: ALL the lead vocals are in the centre channel, which contains more info than any of the other channels. Same mixing philosophy as the RCA 3.0 recordings.

Sanjay

OK, my apologies for this long post, but it's an interesting subject.

The issue, IMHO, is the various ways a mixing engineer and the listener choose to interpret the music. When I go to see a rock band play live, if their PA system is stereo, then if I wander about left to right I will get phantom imaging movement. If their PA is mono, doesn't matter. On the other hand, with the band playing in the rehearsal hall with a small single speaker for the vocalist and everyone else just playing through their amps (and drums natural), the sound/ambiance is quite different, since each instrument is definitely coming from a single source, not a mix of L+R, and thus is nothing like what it sounds over a big PA. What would a 2.0 or 3.0 mix try to reproduce?

For a studio session, where the engineer mixes in a vocal track equally in the L+R, is he/she attempting to create the impression the vocalist is standing there in between the two speakers that are reproducing the rest of the band? Does the listener want that? I agree with Sanjay that if the intent is to provide the sense that the vocalist is right there, using a center-channel only source provides better imaging that a L+R phantom center for the reasons stated: simulating a single source with two sources results in coherence issues.

What about sounds that seem to come from halfway between the center and left (or right)? That is a phantom-offset image that is more tightly bound by use of a center channel, but still can wander a bit, and phase issues as described by Sanjay can occur. It is rare for an instrument to be mixed hard-left or -right. Usually the band is spread across the sound field.

As I've mentioned in a PM to someone involved here, the issue also touches on "what was intended". Just because we've all become accustomed to 2-ch mixes and phantom centers doesn't mean that was the intent of the mixer or artist. Maybe upmixing a 2.0 recording to 3.0 will result in something closer to what was intended.

Finally, a big factor seems to be orchestral music vs pop/rock music. There often seems to be a schism in how those recording are presented. Whether 3.0 helps one vs the other is very opinion-oriented, I perceive.

Using DSOTM as an example, Clare Torrey's vocals on great Gig in the Sky could be mixed as a firm center channel, or could be left as L+R phantom centered. The first case provides a sense that she's right there, the second has her vocals blended in with the rest of the music. What is better? Dunno, but would it have been better if Rick Wright's organ had been mixed more/less centered, or on the same/different side as David Gilmour's guitar? Artistic intent could be a more important aspect with rock music for 3.0 vs 2.0.

Another related issue is this: imagine you could record an orchestra with one perfect microphone per instrument right at the source, all multitracked, and then reproduced with one speaker per instrument, situated appropriately. Does that do the recording justice? No - you miss out on the ambiance of the hall. Unless your living room is equivalent, I suppose.

The "trick" to large orchestral-type "live" music is to reproduce that live sound as accurately and as pleasantly as possible - sometimes those are competing requirements. For rock music, the studio sound is nothing like live, and the end result isn't supposed to sound live usually.

So yes, I think upconverted 3.0 has a place, since it may re-create something closer to the artists/engineers intentions. But this discussion is interesting all on its own.

shinksma

What about sounds that seem to come from halfway between the center and left (or right)?As you said, we still have to rely on phantom images for that part of the soundstage. If the front speakers are spread really far apart, like in a large movie theatre, adding a couple of extra channels/speakers between the L/C/R can help to further stabilize imaging across the soundstage.

This is exactly what Sony does theatrically with their SDDS soundtracks. However, those 7.1 mixes (5 front, 2 surrounds, 1 LFE) are typically used in the the biggest theatres. For the more typical shoe-box auditoriums at your local multiplex, the 5 front channels are downmixed to the standard 3.

Probably overkill for home audio, but I wouldn't mind giving it a try; the extra pair of speakers can be placed closer to my TV, allowing my current L/R speakers to be spread further apart. The more speakers used, the less reliance on phantom imaging. The less reliance on phantom imaging, the more stable the soundstage. I'd be more interested in trying it with music rather than movies. I think upconverted 3.0 has a place, since it may re-create something closer to the artists/engineers intentions.Bringing this back to the original post: from the samples I've heard, the Penteo process only places sounds in the centre channel that would have phantom imaged there anyway. There's no change in soundstage directionality, just greater imaging stability.

Sanjay

OK, sanjay, forgive me for taking a while to get back to this. I was trying to think of exactly how to respond so that it didn't arouse any more of your ire. I am not looking to argue.

Why don't you point out which parts of my explanation are not factual. That would certainly be more productive than attempting to dismiss it by calling it a name.Just because it is "technobabble" doesn't mean, at all, that it is not factual (although some of what you said is sort of theoretical).

That a center channel speaker can anchor the center image is not really arguable. That's overstating the obvious, no?


We aren't discussing what "sounds best", because that is based on personal preference.But that IS what we ARE really discussing. As I said, of course a center speaker can anchor the center image. But the questions are:

Is 3-channel reproduction with the center image extracted and reproduced by a center speaker necessarily the "best" way to reproduce a stereo recording? This answer is very subjective.

Would native 3-channel recordings which utilize a center speaker simply to anchor the center image necessarily be "better" than stereo recordings? Again, I think this answer is very subjective.


As far as "mimicing reality" is concerned, again, what "reality" is it you are trying to mimic? An acoustic trio in your living room? A live rock concert in a stadium? A symphony orchestra in a concert hall? Is "mimicking reality" even truly the goal? I think this differs from situation to situation. There really is no "reality" on an album like Sgt. Pepper's or any other similarly mixed and effected studio album.

I understand the concept of anchoring the center image, but is this really "reality"? If I move from the left side to the right side of the symphony, there really is no static, anchored center image. The "center image" at each location I move to moves with me. If there is an acoustic trio playing in my living room, there is no static anchored center image. As I move around the room, the center image changes. Yes, there is a center image at each location I move to, but it is not static. It changes and is different at every location.

Yes, I understand that the "reality" you are most likely trying to reproduce is that which would occur if you were NOT moving and WERE sitting in the same position. But how do you best reproduce the "reality" of a live situation? Is it best to record "live" with 2 mics? 3 mics? Or can you better mimic a live situation with a multitrack studio recording intended to be replayed on a 2 channel system? A 3 channel system? A 5-channel system? Etc.. The recording methodology used in each of those scenarios would vary depending upon the exact reality one was trying to reproduce.

Suppose you wanted to reproduce the sound of a single acoustic guitarist in your living room. Would it be best to use a single mic and a single speaker? Multiple mics but mixed down for reproduction form a single speaker? One mic replayed in mono form 2 speakers? 3 speaker? Etc.. Why? How about an acoustic guitar duo in your den with a 2-channel recording? Would it necessarily be best to reproduce the sound of each single guitar from each separate speaker or not? Or how about the duo with a 3-channel recording? A trio with a 2-channel or 3-channel recording ? Or how about an amplified acoustic guitar trio in an arena? I'm not really looking for a hard and fast answer to these questions. I'm just trying to illustrate that there are different recording techniques and methodologies that could be used in an effort to provide the desired results.

If you were tasked with having to record a concert at the local arena, for 3-channel playback, how would you do it exactly? Would you attempt to mimic the reality one would hear at a particular location in the arena? How? Would you use 3 mics? If so, where would they be located relative to one another? Or would you use many more than 3 mics and then rely on a studio to mix this recording down to 3 channels? If so, would you ONLY use the center channel speaker to anchor the center image? What about 5 or 7 channel playback? What would be the ideal way to record, live, for 5 channel playback? Again, I'm not looking for any answers. Just providing some hypothetical situations to illustrate that "mimicking reality" is not so easy to define nor accomplish. Exactly how someone chooses to record a source versus how it is intended to be played back and how those two are related is complicated.


I was simply answering your question by explaining what 3-channel recordings do better than 2-channel recordings.I understand what 3-channel recordings can "do" better than 2-channel recordings. That's not really disputable. But do 3-channel recordings necessarily "sound better"? And yes, that is very subjective, but THAT is still the REAL question, subjective or not. I think it is a stretch to broadly say 3-channel recordings are "better" than 2-channel recordings simply because they allow the center image to be anchored. And I think it is even more of a stretch to say that 3-channel reproduction of 2-channel recordings is "better" simply because it anchors the center image.


The answer is in the recent 'Love' DVD-A, which he and his son Giles mixed: ALL the lead vocals are in the centre channel, which contains more info than any of the other channels.Yes, exactly, the center channel was utilized discretely. Placing the lead vocals in the center is a reasonable and obvious choice, same as is often done with a 2-channel recording. But is placing the voices there, discretely, really "mimicking reality". I suspect that there are also cues in the L/R (or even surround) channels with most of the lead vocals so as to provide some of the same effects that a 2-channel recording could provide with the vocals. But the 3rd center channel can and is used for more than simply anchoring a center image just as the surround channels are used for much more than providing ambiance. And that was my point. A 3-channel recording may best a 2-channel recording because of the new possibilities it provides. Anchoring the center image being one of them, but it can certainly be used for much more than that with a studio recording such as these, where there really is no "reality". As I said very clearly, specifically 3-channel recordings would interest me, particularly studio recordings, not simply because the center channel image can be anchored, but because of all the new possibilities having a 3rd speaker provides. Same for 5-channel studio recordings which, for most, "mimicking reality" is not the goal at all.


Same mixing philosophy as the RCA 3.0 recordings.But there really is no mixing philosophy, here. They are simply recorded live with 3 mics. Where they placed the mics was, basically, the only mixing philosophy involved, really. But how and why they placed the mics where they placed them would be interesting to know. One would think that with these recordings they were trying to "mimic reality" but would this be how one would set about recording an orchestra, live, for 3-channel playback, today, whether one was limited to 3 mics or not?



Now, to get back to Penteo's question, NO, I would not like my stereo recordings to be reproduced with the center image (and does this mean ALL mono info?) extracted and reproduced by a center channel speaker. You, I think, would. Different strokes. As you said, this is very subjective. There is no "correct" answer.

Well this certainly is a lively discussion... thanks for taking this discussion of panorama placement so seriously, guys, I wish everyone did.

What we're trying to recreate with Penteo (especially in 3.0 mode) is not exactly REALITY; but it is trying as hard as we can to convey the same experience that the mix engineer had when they were mixing their work of art in stereo. Because they were sitting in the sweet spot, it would have sounded to them as if they had a center speaker. Any of us who have worked as mix engineers can tell you that in a stereo mix session, especially in a large, expertly designed and engineered control room, it always, always, always, always - did I say always? sounded like three speakers instead of two. The studio lighting is/was usually set up so that you can't see the speakers, so a lead vocalist's voice was always a firm, solid, point source, right dead between the speakers. (Of course I'm talking about main speakers in a soffit; not nearfields.)

In a properly engineered mix room, you can stare directly at the left or right channel speakers and hear absolutely nothing coming out of them... when there is a full-level vocalist coming from directly between them.

Having a 3rd (center) speaker simply gives the material that happens to be in-phase and in common amplitude a home that doesn't wander. In a live performance, that could be anything from a single violin player to the center section of an entire orchestra. Of course in pop music, it's usually the bass and lead vocal. One is acoustically centered, the other is artificially placed there by a pan pot. In either case, it was undoubtedly the intention of the recordist and/or mixer to have that sound come from between the speakers, from a phantom speaker. Anything else would pretty much go against all of the concepts of phase alignment, recording, and acoustics.

If you are sitting in the sweet spot between two speakers, as you are behind a recording console, you can't/shouldn't tell the difference between 3 speakers and 2 speakers. If there is, then there's a phase problem somewhere. There has a tremendous amount of engineering science that has gone into making left and right channels as symmetrical as possible, all the way to having studio monitors that have tweeters placed as mirror images of each other, all just to make the two sound point sources sound like three (or four or five).

Problem is, it only works if you're sitting in the sweet spot. The only way to have it stay in place for more than one listener is to have three sound sources, otherwise the center image can't help but wander around as you do, since your brain/ears no longer have any valid point-source information, since your brain/ears can't determine common phase and amplitude when you're not sitting perfectly between the speakers.

=John

Anything else would pretty much go against all of the concepts of phase alignment, recording, and acoustics.
Thats incorrect.

With "focused panning", where panning to the center sends the sound to the center channel rather than to the middle of the soundfield, this technique usually doesn’t work too well for music, music usually requires divergence panning. Then on top of that theres (VSP)Virtual Surround Panning techniques that are used in the pannning process which contain early reflections, correlated relative direction and delayed timing characteristics. A 3.0 mix does not simply have to be (3) monophonic signals.. or simple grouped adjacent stereo pairs.

I was trying to think of exactly how to respond so that it didn't arouse any more of your ire.What "ire"? I take the time carefully explain my points and you see that as anger? As far as "mimicing[sic] reality" is concerned, again, what "reality" is it you are trying to mimic?It doesn't matter what "reality" the recording engineer and artist intend to create, what's more important is how it is conveyed to listener(s).

If their reality means an instrument solo is at the centre of the soundstage, then that's where it should be, not collapsing to the nearest speaker whenever the listener moves from the sweet spot. If their reality includes a vocalist at the centre, then it should sound like the human voice they recorded, not some dual-mono phantom-image coming from two locations.

Using a centre speaker doesn't take anything away from the L/R speakers. Those speakers are still there. The centre can only add, by helping to overcome limitations of 2-speaker playback. If I move from the left side to the right side of the symphony, there really is no static, anchored center image. The "center image" at each location I move to moves with me.If there are instruments at the centre of the orchestra, do they travel along with you when you move from left to right? Or do they stay at the same location in relation to the orchestra?

If those instruments don't travel with you but instead stay at the centre of the orchestra, then how are you going to reproduce that using only 2 speakers? I think it is a stretch to broadly say 3-channel recordings are "better" than 2-channel recordings simply because they allow the center image to be anchored. And I think it is even more of a stretch to say that 3-channel reproduction of 2-channel recordings is "better" simply because it anchors the center image.Again, "better" is subjective, based on personal preference. A single speaker will reproduce the human voice more like the way it is produced in real life than two speakers can (no one talks in dual-mono). That doesn't guarantee that you will like it better, just that it comes closer to mimicking reality. But is placing the voices there, discretely, really "mimicking reality".Yes, unless you know someone whose voice is a phantom image?

Sanjay

Thats incorrect.

With "focused panning", where panning to the center sends the sound to the center channel rather than to the middle of the soundfield, this technique usually doesn’t work too well for music, music usually requires divergence panning. Then on top of that theres (VSP)Virtual Surround Panning techniques that are used in the pannning process which contain early reflections, correlated relative direction and delayed timing characteristics.

Dude, I really, really do appreciate your perspective, because you are SO on the opposite side of the spectrum from me. I'm a third generation sound engineer, and somehow I feel like I have a biased perspective on this. One thing that this thread has shown me is that there are two camps:

Camp #1: Old school. Those of us (like me) who have been around since the days of mono. Most of us are here in the Bay Area now, working in or around film sound. Some of us have shelves full of awards because we spent half of our lives in control rooms that were the size of a new car showroom, angled ceilings and walls that were tuned for no resonant points. They were expertly tuned and time-aligned, but with only two, very large full-range monitors, usually JBLs or Westlakes, (of course we always had Auratones(!)) that were built into a soffit above the window into the studio. Symmetry between left and right was always critical, so that the virtual 3rd speaker would be as prominent as possible. We spent half of our lives trying to make it sound like the glass wasn't there, and have invisible speakers, since what we were attempting to recreate was the experience of being on the other side of the glass; so that if you listened on headphones at home, it was like being a VIP or a celebrity who had the opportunity for a private audience in an acoustically engineered studio with the singers and musicians.

Camp #2: New school. Electronic music, often not even originating from acoustic sources, often completely synthetically immersed in a virtual soundfield. That's when we start using "divergence panning" and "virtual surround panning" techniques; those would tend to reproduce a rock concert environment or some sort of "virtual reality". Any acoustic recordings can often take place in untuned, random rooms, since microphone placement is close -- the actual environment will be generated by an algorithm. Don't get me wrong, I'm not saying that this is any less of an artistic format.

It's funny, I've actually had this discussion -- in discussing Penteo -- with Rob Jaczko, Chairman of the Music Production and Engineering Dept. at Berklee College of Music in Boston. Pretty much the conclusion is:

* The recordings of pre-1985 or so are designed to reflect recording studio accuracy as much as possible, using as little as possible between the microphone and the loudspeaker, often with substantial distances (sometimes several feet) between a musical instrument and the microphone in order to capture the ambience of the studio.
* Post 1985 or so are much more abstract works of art created by electronics, in which the ambience is generated by an algorithm.

((One nice thing about Penteo upmixing is that it doesn't care; all it's doing is coorelation and placement that is based on the original panning positions, whether those positions were done by a human or by a machine doesn't matter. If there are phase/delay issues, then they are replicated in a 5.1 field. But no divergence panning.))

Divergence panning. Great term. That means, don't pan anything into the center, except for special effect. Keep the lead vocal in the left and right speakers so that it doesn't "focus". It's funny, because that's based on art replicating art attempting to replicate reality. It's an attempt in 5.1 to reproduce stereo, which in turn is trying unsuccessfully to replicate mono, since there's no center speaker. But in 5.1, there is a center speaker, so we keep it out of the center to make it sound like stereo recreating mono. Confusing. My brain is smoking.

So many people -- 40 and under -- have now grown up with mono meaning "coming out of two speakers simultaneously" that they think that that's what mono means. Mono means (and this is in the Phil Spector sense, present newsworthy items not withstanding) a single point source. One speaker. Mono.

We can all agree that human voices do not come normally from two simultaneous places, ten feet apart, in the front right and front left corners of a room.
But -- AMPLIFIED human voices -- over stereo sound systems -- DO come from two simultaneous places. At some point the art of music started replicating the art of stereo music attempting to replicate a single point source, unsuccessfully.

With "old-school" recordings, Beatles, Stones, et. al; there really was an attempt to create a virtual 3rd speaker; there always was. It was just that when it got home, there were only two speakers, not three, so when you got it home, the voice of, say, Elton John, was coming from two speakers simultaneously. So many people at home now find it uncomfortable to hear a mix the way the mixer actually heard it -- that is, with a 3rd speaker. In some control rooms, it was even real, in some, it was virtual. (Almost all 3-channel rooms were in Nashville.)

But it was ALWAYS focused. You mix, staring at that sole point-source voice coming at you from dead-center, wishing there were three channels to take it home in. Mixers have done that for decades.

This is heavy.
Your thoughts?

PenteoSurround very succinct post. Since we are discussing art, overall concepts of point sources, contrived virtual soundfields, accuracy aka reality is entirely subjective. Trying to express the sublime is NOT science.
At some point the art of music started replicating the art of stereo music attempting to replicate a single point source, unsuccessfully.Panning instruments is not just simply placing point sources at or between speaker locations. But it was ALWAYS focused. You mix, staring at that sole point-source voice coming at you from dead-center, wishing there were three channels to take it home in. Mixers have done that for decades.Depending on ones mixing (aritst & engineers) philosophy, voices could actually originate from any direction, even from behind.

We can all agree that human voices do not come normally from two simultaneous places, ten feet apart, in the front right and front left corners of a room.No. But human voices (or acoustic instruments) don't really, once they reach our ears, entirely come from a single point source either (unless you're in an anechoic chamber). The room does affect, sometimes to a great degree, the perceived sound of that voice (or instrument). Relying upon (or expecting) the listener's own room, upon playback, to provide the particularly desired cues and ambiance of a voice (or instrument) that simply originates from a single speaker at a fixed distance would not really make any sense, would it? What makes sense is mixing the desired ambiance into the R/L speakers (whether a 3rd center speaker is used or not) and relying (albeit hopefully) on the listener's ability to dictate, control, and manipulate their own playback environment so as to best reproduce the recording as intended.

human voices (or acoustic instruments) don't really, once they reach our ears, entirely come from a single point source either (unless you're in an anechoic chamber). The room does affect, sometimes to a great degree, the perceived sound of that voice (or instrument).The human voice does come from a single point. It is reflections that come from all over the room. Let's not conflate one with the other. Relying upon (or expecting) the listener's own room, upon playback, to provide the particularly desired cues and ambiance of a voice (or instrument) that simply originates from a single speaker at a fixed distance would not really make any sense, would it?Sure it would. How is a singer at the centre of the soundstage and a speaker at that same location different? Both will be the same distance from room boundries and will exhibit reflections that correlate well. Compare that to splitting that voice and having it come from two speakers, both of which are closer room boundries than the singer, meaning they will generate different ambience cues (especially sidewall first reflections) than the singer would.

Sanjay

human voices (or acoustic instruments) don't really, once they reach our ears, entirely come from a single point source either (unless you're in an anechoic chamber). The room does affect, sometimes to a great degree, the perceived sound of that voice (or instrument).The human voice does come from a single point. It is reflections that come from all over the room. Let's not conflate one with the other.And is that not EXACTLY what my post implies? :confused:


Relying upon (or expecting) the listener's own room, upon playback, to provide the particularly desired cues and ambiance of a voice (or instrument) that simply originates from a single speaker at a fixed distance would not really make any sense, would it?Sure it would. How is a singer at the centre of the soundstage and a speaker at that same location different? Both will be the same distance from room boundries and will exhibit reflections that correlate well.How can my small carpeted and draped (and far from ideal) den possibly and/or appropriately allow for the proper "mimicking" of the sound of a voice or instrument created at a much greater distance in, for example, a large concert hall, a church, an arena, outdoors, etc.? Is the goal recreating the sound as it would sound in my den or as it would sound in another environment altogether? Could be either one, I suppose, but most (no, not all) of the recordings I encounter are not trying to recreate the "reality" of my den. In fact, many studio recordings are not interested in recreating the cacophony as it would occur in my den at all, but instead are geared toward reproducing a more transportive and imaginary soundscape than that.

Yes, the sound of a voice originating in my very own den, from exactly the same place that my center speaker is located, is probably best reproduced by my center channel speaker, alone. But, honestly, I'm not entirely certain of that. The projection of a human voice is completely different than that of a single speaker. I DO realize that certain ambient effects do not necessarily rely upon 2 (or more) speakers and can be effectively mixed into a single channel.

Compare that to splitting that voice and having it come from two speakers, both of which are closer room boundries than the singer, meaning they will generate different ambience cues (especially sidewall first reflections) than the singer would.I realize that splitting it between 2 speakers is not ideal, either. But neither is simply reproducing the sound solely from the center channel speaker. Given the choice, in most instances, I would probably take the possibilities afforded by the 2 speakers over the limitations of a single speaker. But as I have alluded to, with many recordings, if a 3rd center speaker were/is meant to be utilized, it is very most likely that at least SOME of the environmental ambience cues of the center channel "image" WOULD be encoded in the L/R channels.

As I also said, but you omitted:
What makes sense is mixing the desired ambiance into the R/L speakers (whether a 3rd center speaker is used or not) and relying (albeit hopefully) on the listener's ability to dictate, control, and manipulate their own playback environment so as to best reproduce the recording as intended.I think that this is the philosophy used with most recordings. To rely upon the room to provide the proper effect upon reproduction is wishful thinking given the large variety of rooms most recordings are reproduced in. The whole purpose of room treatments is to try and remove the room's undesirable and variable effects from the equation.

And is that not EXACTLY what my post implies?No. You said "human voices (or acoustic instruments) don't really, once they reach our ears, entirely come from a single point source either". In fact the human voice does reach our ear completely from a single point. Reflections arrive later, after we've heard the voice. I made the distinction because we're talking about localizing the source, not its reflections. How can my small carpeted and draped (and far from ideal) den possibly and/or appropriately allow for the proper "mimicking" of the sound of a voice or instrument created at a much greater distance in, for example, a large concert hall, a church, an arena, outdoors, etc.?Who said one speaker would do all that? If the recording is trying to convince you that there is a singer at the centre of the soundstage, pressurizing the air in front of you, which playback method will come closer to doing that: a speaker pressurizing the air in front of you or two speakers pressurizing the air at your left and right?

Voices aren't isolated in the centre channel. There's much more in there, including ambient cues. That's why, even when you're listening to the tinny mono speaker on a clock radio, you can still tell whether the singer is in a reverberant hall or an acoustically dead space.

Keep in mind that we're not talking about 2 speakers vs 1, we're talking about 2 speakers vs 3. Just because you use a centre speaker for centre imaged mono content doesn't mean you stop using the L/R speakers for all the stereo information that images elsewhere in the soundstage. The projection of a human voice is completely different than that of a single speaker.How so? Is a dual-mono phantom image closer to how a human voice is projected compared to a point source?

Sanjay

No. You said "human voices (or acoustic instruments) don't really, once they reach our ears, entirely come from a single point source either". And that is fact! Sanjay its apparent you are unfortunately conflating "playback methods" with mixing philosophy or paradigms.

Most joystick panners have a width or focus control that determines the proportion of a track that is routed to the physical center and phantom center, the engineer can decide whether to place the listener in front of the band, in the middle of the stage, or in some other strange place.

Btw mixing for film release is NOT the same as music production i.e. a vocal is NOT usually isolated (focused panned) exclusively to a center channel. Period.

If you can't accept this "reality", then by all means please visit every recording studio asap, delete all associated algorithms and while your at it rip-off the physical hardware to...God forbid we allow such madness as VSP in the first place. Thanks in advance:)

And that is fact!Nope. The reason we can localize a vocalist at the centre of the soundstage instead of all around us is because we hear the difference between a human voice and its room reflections. The voice comes from a single point, the reflections come from all around. That's not playback methods vs mixing paradigms, it's plain physics. Btw mixing for film release is NOT the same as music production i.e. a vocal is NOT usually isolated (focused panned) exclusively to a center channel. Period.There are DVD-A titles (e.g., the aforementioned 'Love' album) and SACDs (e.g., James Taylor's 'JT') that have lead vocals mixed exclusively into in the centre channel. Apparently established music mixers like George & Giles Martin and Nat Kunkel don't have a problem loading centre imaged content into the discrete centre speaker, just like a film soundtrack is mixed. "Period" indeed.

Sanjay

Sanjay you’re apparently having significant difficulty following the evolution of the debated subject matter. Telling an artist or an engineer how to paint or mix is laughable.:p

This whole exchange reminds me of a documentary that aired on PBS regarding a artist who was criticized most of his career, stating that his paintings were child-like in execution and his technique was lacking. In fact everything about his work was WRONG! Now his work hangs in some of the most respected galleries in the world.

So Sanjay... please do not come within ten feet of any mixing console. Thanks in advance!

Sanjay you’re apparently having significant difficulty following the evolution of the debated subject matter. Telling an artist or an engineer how to paint or mix is laughable.:pExcept he's NOT doing anything of the kind. His explanation of the distinction between the actual voice (single point source, with the mouth pressurizing the air at that source) and the reflections from various room surfaces (which ALL arrive at a different time from the original single point source to the ears) is EXACTLY correct. Whether or not you prefer the mixer to try to replicate THAT PARTICULAR situation is irrelevant to his explanation. He is also quite correct that if you start to wander around a room with a live performer(s) in front of you, IN REAL LIFE, what is emanating from the centre of the stage remains anchored there whereas a 2 channel reproduction via phantom centre does NOT remain anchored in the centre. The one you prefer is entirely your choice, but the one that anchors the point source is closer to reality.

As for "non-reality" mixes (as outlined by PenteoSurround above--mostly as post 1985 mixes, to be more specific in the reference), there may well be no distinct "centre" source point--but I doubt Sanjay is arguing that one NEEDS to create it to be a "legitimate" mix. His point is quite clear (and I sense from a few responses to him in this thread that this is not the first time some have argued with him)--IF one wants to more accurately recreate a single point mono source in the centre of the soundstage, then a hard centre is better FOR THAT PURPOSE than a phantom centre. That in no way implies that one should always seek to re-create such a situation, NOR does it imply that one NEEDS to prefer that kind of re-creation. It DOES explicitly state (and quite correctly) that for that specific purpose, a hard centre is the better choice--JUDGED ON THE SINGLE CRITERION OF RE-CREATING A SINGLE POINT MONO SOURCE AT THE CENTRE OF THE SOUNDSTAGE. It really isn't all that hard to understand, though it appears some have an agenda with the goal of obfuscating his point (it seems routed in the notion that because Sanjay prefers MCH playback, matrixed as well as discrete, that he is offending the sensibilities of some who prefer 2 channel playback UNLESS it is discrete (if at all)).

As to mixing "like the movies", again, you are misconstruing his point. He is not saying all mixers SHOULD mix like for movies but that using that approach CAN be successful with music. The distinction is clear and important.

IF one wants to more accurately recreate a single point mono source in the centre of the soundstage, then a hard centre is better FOR THAT PURPOSE than a phantom centre. I never said anything about creating a "single point source", read my postings again. I'm not making up anything new to the pro mixing community. It's not all or nothing...btw I'm referring to the middle of the soundfield..a real dimensional space that Sanjay has yet to comprehend.

I never said anything about creating a "single point source", read my postings again. I'm not making up anything new to the pro mixing community. It's not all or nothing...btw I'm referring to the middle of the soundfield..a real dimensional space that Sanjay has yet to comprehend.

Sorry to butt in on the conversation, and I'm not specifically trying to defend Sanjay...

I don't think Sanjay has failed to comprehend the "dimensional space" aspect of the center of the soundfield. I think Sanjay is just pointing out that if one wanted, you could isolate the component that is exactly centered as a mono track from the center speaker of a 3.0 system. The ambiance portion of the audio that is not exactly centered (reflections, real or simulated, other effects) would have to remain in the stereo L+R pair. If one didn't want to that, that's OK. But some mixers do prefer to do that, making a very dry-sounding center channel if isolated from the L and R pair.

I think the point is being overemphasized on both sides, and that is causing some friction. Or maybe you guys always sound like this...

Sanjay's position is that, from a pure "engineering" point of view, to reproduce a vocal track such that if you shift your position or turn your head it still sounds like it comes from the exact front center, you need to isolate as much as possible of the direct (dry) vocal to that center channel. Use of the L and R would be to re-create whatever "chamber" acoustics or other funny effects are desired.

On the other hand, tbrunet's position is that the mixer can do whatever he/she feels like to achieve the desired ambiance and directional effect.

Neither of these positions conflict, they just represent the two parts of the science and art of mixing, IMHO.

Which boils down to the original issue: if the original artist and/or mixing engineer really wanted to produce an anchored center vocal but couldn't actually do that due to the limitation of 2-channel media, then I would understand completely if a 3.0 mix was created and released.

tbrunet, when you speak/type of the dimensional space of the phantom/anchored center, what do you mean? I see it as a single linear dimension that is being discussed, since anything that is not present equally in both L and R is part of the stereo image must remain in the L+R pair. So it is a case of how much goes equally into the L+R to create the phantom center, and how much goes into the center as a hard center. So there is only one dimension to vary. Unless the tonal character of the fixed C is different from the L+R phantom, in which case I suppose that can create some depth as you move off-center/turn your head.

IMHO, and maybe I'm missing the issue that is really underlaying the lively discussion...

shinksma

Telling an artist or an engineer how to paint or mix is laughable.Who's doing that? I simply provided examples of multi-channel music (from well established mixers no less) which directly contradict your claim that vocals are not focus panned exclusively to the centre.

Sanjay

it appears some have an agenda with the goal of obfuscating his pointYou'll have to forgive my fan club. If you check tburnet's posting history, you'll see he follows me from thread to thread to thread attempting to misrepresent my points. It never works, which is why he always resorts to ad hominem comments about me.

Sanjay

Btw mixing for film release is NOT the same as music production i.e. a vocal is NOT usually isolated (focused panned) exclusively to a center channel.
...vocals are not focus panned exclusively
Try again Sanjay! Usually does not have the same meaning as NOT or always:)

Usually does not have the same meaning as NOT or alwaysWhich is why I didn't use the word "always". But they are examples nonetheless of recording engineers placing centre imaged sounds in the centre speaker. Those recordings do in the mix what the Penteo process does in post.

Sanjay

Who's doing that? I simply provided examples of multi-channel music (from well established mixers no less) which directly contradict your claim that vocals are not focus panned exclusively to the centre. Again!

I DID NOT SAY THAT VOCALS ARE NOT FOCUS PANNED

The subject of mixing virtual soundfields were being addressed, and I stated a fact that musical instruments including the voice are not USUALLY mixed as a simple point source and or with simple amplitude panning.

The subject of mixing virtual soundfields were being addressedOnly by you. This thread started off addressing the upmixing of 2-channel content to 3 channels by extracting correlated mono info from the L/R channels, and that's what the rest of us are discussing. If you want to change the discussion to address virtual soundfields, that's fine. But it has no bearing on the Penteo process nor upmixing 2 channels to 3 channels.

Sanjay

Dude, I really, really do appreciate your perspective...

This is heavy.
Your thoughts?;)
If you want to change the discussion to address virtual soundfields, that's fine.No one is changing the discussion!

No one is changing the discussion!You didn't see his orignal reply to you.

So if it is on topic, what does virtual soundfields have to do with Penteo or any other centre extraction process?

Sanjay

tbrunet, when you speak/type of the dimensional space of the phantom/anchored center, what do you mean?
http://books.google.com/books?id=ZgdkVhHUjEMC&pg=PA51&lpg=PA51&dq=wide+stero+phanom+image&source=web&ots=jZreGtDjHh&sig=qFxGP53mpTKU1nokQLMnF1M3EO8&hl=en#PPA52,M1

“Its possible to generate a spread image that appears to occupy an area. It’s a phantom image that has a “size” that extends between two audible boundaries. The potential size of the spread image is perceptually variable; it might be slightly wider than a point source or it may occupy the entire stereo array. A defined spread image can also be panned across the soundstage while it maintains it perceived width.

The lead vocal in the Beatles album (Abbey Roads) “You Never Give Me Your Money” begins with lead vocal as a point source, then the image soon becomes a spread image that gradually grows wider and ultimately occupying a significant amount of the sound stage.

Two important distance cues which shape recorded music:

1) The distance of the listener to the sound stage.
2) The distance of each sound source from the listener

The perceived performance environment establishes the front edge of the sound stage with respect to the listener and determines the level of intimacy for the production."

Pick up the RCA Living Stereo and Mercury SACDs. Yes.
John, do you have any other selections in addition to "Help Me" from Joni Mitchell you can put on your website.

Pick up the RCA Living Stereo and Mercury SACDs. Yes.
John, do you have any other selections in addition to "Help Me" from Joni Mitchell you can put on your website.

I'm not sure I'm keeping that one! It's very difficult to put anything on the web that is not going to infringe on someone's copyright.

Getting back to the topic: There are definitely two camps, mostly determined by the age of the listener. It all has to do with the loudness of guitar amps that started in the rock era:

Almost anyone under the age of 40 has probably never heard a professional musical performance that wasn't amplified. It HAD to be, because the guitar amps were so loud. So in that case, their sense of musical reality isn't that of listening to a singer standing 20 feet in front of them, emanating from their vocal cords and mouths; it's coming from the amplified speakers that are reproducing a live stereo mix, flanking the band on both sides of the stage.
People who are 40, 50, and older can usually remember being in a much more intimate performing space, in which they were listening to the actual singer's voice coming from the singer's mouth, 20 feet in front of them, singing with unamplified acoustic guitars, piano, upright bass, etc.

For that reason, and specifically because of the feedback that I have gotten from this thread, I have decided that Penteo will always offer three mix options, similar to DPLII's "Music Mode" vs. "Movie Mode".


FOCUSED CENTER mode. This is equivalent to DPLII's "Movie Mode", in which the center channel material only emanates from the center speaker. The sides, both front and rear, carry the material panned off to the appropriate side. This is for fans of #2, above, who like having the intimate feel of a vocalist or a musician focused directly in front of them. It's also used in motion picture upmixes, in which the dialogue comes from the center speaker only.
DIFFUSED CENTER mode. This is more like DPLII's "Music Mode", but in the case of Penteo, it's simply the original stereo mix in the front left and front right channels, with the extracted pure center in the center, and the extracted pure left and right in the rears. Everything is perfectly in-phase (none of that 90-degree foolishness) so the left rear and center form a phantom "left center" in front of the left front speaker, so that the 10AM position is centered in front of the left front; likewise the right rear and center form a phantom "right center" in front of the right front speaker, so that the 2PM position is centered in front of the right front. So the fronts and center acoustically steer everything toward the center performance space, but if you walk by or listen closely to the rear speakers, you are treated to isolated performances that were originally panned hard right or hard left. This is for the fans of #1 above, who prefer having the lead performer coming from a less-specific point in space.
3-FRONTS-ONLY mode. This isn't surround at all, the rears are silent. It is the exact equivalent of a stereo mix, but with the phantom center being replaced by a focused center coming from the front speaker. It's identical in level, in every respect to the original stereo mix, otherwise. This is for stereo purists who want to hear what it would have sounded like in the sweet spot of the recording console, where phantom centers sound exactly like hard centers. But it allows you to be somewhat off-center, or have a less-than-perfectly-symmetrical listening environment, and still hear the phantom center portion coming from the center.

In ALL the above cases, the mix, when downmixed back to stereo, EXACTLY equals the original stereo; it is always tested against the original stereo mix 180 degrees out of phase to assure cancellation -- proving a perfect recreation of the original mix.

It all reminds me of one of my favorite phrases:
"That's why they make chocolate AND vanilla". Oh yeah, and chocolate chip cookie dough.

I sincerely thank you all for your feedback in helping me to make these decisions.

Looking forward to some releases.