Some noobish 5.1 panning questions

[iguana]

I'm a composer by nature and still somewhat of a novice when it comes to post production, but it's a topic I find rather intriguing. I've provided 5.1 DVD mixes for a few short films, and I like to think that they've turned out okay, but of course, there's a lot of room for improvement. So, I'd love to get some advice about a couple of issues that I haven't quite wrapped my head around yet.

Generally, I'm talking in the context of a traditional 5.1 movie mix, with dialogue in the Center, most of the action happening in the front, etc.:

- When you're panning a mono source from left to right in the front (or you're just placing it statically somewhere inbetween), how do you treat the Center channel? Do you strictly pan from L to C, then from C to R, thus making the Center channel absolutely equal to L and R? Or do you do an ordinary L-R pan, with a portion of the signal fed to the Center as an additional anchor? Or do you leave the Center out altogether?

- Generally, how do you make a stereo source fit into LCR? Say you have a stereo ambience that you'd like to place in the front. Do you just put it on LR and keep it out of the Center, or do you feed a mono sum into the Center as well?

- What if the stereo source needs to be condensed and/or moved across the front, such as for a vehicle that takes up some space on the screen? Do you treat it as two mono sources and do a traditional LCR panning with a "fixed" distance between them, or do you just fold it down to mono no matter what?

- When you do a F/B-panning, are you going out of your way to minimize phase correlation between the front and the back signals? I've first read about the HRTF issues that result from having similar signals on the front and back at the same time in Tom Holman's book, and I can definitely reproduce the effect – I'm just not sure how much it really matters. For instance, if it's a serious issue, why do so many post consoles have surround joysticks – I would think that these practically invite the problem?

Any insights about these topics are most welcome, as they've been keeping me awake at night for quite some time

[DynaForte]

Originally Posted by iguana

When you're panning a mono source from left to right in the front (or you're just placing it statically somewhere inbetween), how do you treat the Center channel? Do you strictly pan from L to C, then from C to R, thus making the Center channel absolutely equal to L and R? Or do you do an ordinary L-R pan, with a portion of the signal fed to the Center as an additional anchor? Or do you leave the Center out altogether?

Leave it out altogether. Treat the center as it's own world, the LFE as it's own world - and the four others as their own three dimensional spacial world.

Originally Posted by iguana

- Generally, how do you make a stereo source fit into LCR? Say you have a stereo ambience that you'd like to place in the front. Do you just put it on LR and keep it out of the Center, or do you feed a mono sum into the Center as well?

Same as above. Use stereo straight to LR. Of course for some ambiences you can do one that's Left back, left front for a side feeling, or right back right front - or back left back right etc...

But if in front it's L and R

Originally Posted by iguana

- What if the stereo source needs to be condensed and/or moved across the front, such as for a vehicle that takes up some space on the screen? Do you treat it as two mono sources and do a traditional LCR panning with a "fixed" distance between them, or do you just fold it down to mono no matter what?

As far as panning, pan with L/R not LCR.
As far as "condensing", If I understand you right, you can pan them both to the "phantom center", but not actually put them into the center channel.

Originally Posted by iguana

- When you do a F/B-panning, are you going out of your way to minimize phase correlation between the front and the back signals? I've first read about the HRTF issues that result from having similar signals on the front and back at the same time in Tom Holman's book, and I can definitely reproduce the effect – I'm just not sure how much it really matters. For instance, if it's a serious issue, why do so many post consoles have surround joysticks – I would think that these practically invite the problem?

I'm not totally sure what you mean.

[georgia]

when I do placement in a sound field and work sounds in a mix I pan some sounds L C R sometimes L R leaving C feed off. Some times I do L C R with just a small percentage of sound going to C. Sometimes I have the same sound in all Front speakers and just do a bit of percentage work to balance where I want it. When I start to move material from front to back Some times I sweep from L to Ls only, sometimes R to Rs only, some times I take a stereo sound that is L and R only and pan past the listener to Ls and Rs only. Sometimes I take point sources and "fly" them around through the room along the "walls" L to C to R to Rs as an example...100 percent hard over if you will. Some times it's right down the center Ls and Rs paned from the read to C mono... Look it's all a matter of taste and requirements. There are many ways to move sound around a room, many ways to affect a simple pan. When it comes down to it, the most important issue in panning is to do what works best for the picture and the story first. Don't get cute for the sake of cute. But don't be afraid of a joy stick or two. Be aware of phase correlation issues when you mix but don't let it scare you into submission.... Just be careful. Keep an eye on phase, and levels. When you finish your passes and you are happy with your 5.1 or 6.1 mix. check it in LCRS, LtRt encoded and decoded monitoring, Stereo and Mono to make sure its' still working adequately and you are not creating any downmix issues. Remember so folks will listen on a cheap TV set others will be sitting in a 3D or IMAX Dolby or DTS 6.1 theater, some will hear it on an optical LtRt mix....

just make sure it sound good, works well, and does what it's supposed to do, ...enhance the story.... and make sure you meet delivery specs.

Enjoy and have fun

cheers
geo

[iguana]

Originally Posted by DynaForte

As far as panning, pan with L/R not LCR.
As far as "condensing", If I understand you right, you can pan them both to the "phantom center", but not actually put them into the center channel.

Interesting, does that mean you keep everything out of the Center except for the dialogue? Or do you do use it for other stuff (foley etc.), just never in combination with L and R?

I'm not totally sure what you mean.

I'm talking about the fact that if you put the same signal (or signals that are similar, phase-wise) on the front and back simultaneously, you get these weird phasing effects when you move your head. Tom Holman recommends to use something that decorrelates the signals (such as a -5 cent pitch shifter) on the rear channels whenever you do a dynamic front/back-pan. I'd like to know whether people follow this practice in professional mixes, or if they just don't bother (for instance, I'm not sure if a cinema with its speaker arrays would alleviate the problem or make it worse).

Georgia, thanks for your answer! It puts my mind at ease that there doesn't seem to be a horrendously wrong way to tackle this problem. Of course, I'm always looking for best practices, not at least because I want to avoid the worst pitfalls of creating utterly non-translating mixes without having the budget for a proper post mixing environment.

[danijel]

Originally Posted by DynaForte

Leave it out altogether. Treat the center as it's own world, the LFE as it's own world - and the four others as their own three dimensional spacial world.

Not questioning if this works for you, but you make it sound as though it's a hard and fast rule.

If I want to pan a hard FX slightly to the left, most of the sound will go to the C, a little to the L, and, if it is an interior, reverb will go everywhere.

If the dialogue is strictly centered, then you will probably want all the foley in the center as well (at least the sounds that characters make themselves).

As Geo put it, don't worry too much, there are a lot of ways to do it right if you follow your ears. Most people pan car-bys across LCR as mono sources (because its the easiest way), but if you have a car entering the shot close-up, go for panning the L and R source the way you imagined it. That can sound awesome - I always do it with approaching cars - they come from C, and then spread out to LR (but leaving a lot in C, for anchoring).

There are two posts by user ggegan about panning dialogue and fx that were really revealing to me, but who knows where they are now Try and find them!

[ErikG]

I second what Danjiel wrote.
There ARE no real rules to go by. It's all about the film and you will have to deal with the exact same type of scene differently each time as it will need to be handled differently. It is also affected on how the scene was cut (sound editing and picture wise), who cut the sound, and how the mono and stereo sounds position in the 5.1 space.

My experience is that sounds that are primarily lower frequencies can be panned wider than mid rangey sounds. Sounds with a norrow frequency range will be harder to pin-point. And wide range sounds might increase mushyness, and be pulling the sound center away from the center of the screen.

So the answer is, make it work in a calibrated room, check it (as already mentioned) in all delivery formats, Personally I still check for mono and stereo fold down compatibility, but I seldom trash a good 5.1 sound just because it wont work as well in any other format, but I will, time permitting do alt mixes/panning for important sound cues for LtRt and stereo mixes.

Practice makes... ok perhaps almost never perfect but at least you learn on the way.

I have yet to make a mix I was perfecty happy with. So I try to keep on learning, as I have for the past 19 years (jeez how time flies!).

[Key]

I would agree with DynaForte personally. Just because I think this will work best in most cases. If you pan L>C>R and then play that back on a consumer system with a mismatched center you can get a weird shift that wont be there in most cases if you just go from L>R.

But I am kind of anti center channel personally. I think at best when you have a screen dead center in between L and R the center acts more like a height channel.

[georgia]

Originally Posted by iguana

Georgia, thanks for your answer! It puts my mind at ease that there doesn't seem to be a horrendously wrong way to tackle this problem. Of course, I'm always looking for best practices, not at least because I want to avoid the worst pitfalls of creating utterly non-translating mixes without having the budget for a proper post mixing environment.

hey, sometimes I do stuff with L and C or C and R. sometimes I've been know to build dynamic ambiances using various stereo sounds across all sorts of pairs, L-R, L-Ls, R-Rs, L-Rs, R-Ls, L-C-R-Ls-Rs, then I might very slowly move the various ambinces to build a bit of "movement" in a shot... but then again, most of this work is probably never noticed by most clients, but I like to build dynamic mixes. I tend to disagree with not utilizing the center channel. I'll drop dialogue and spot mono sfx, like Foley for example in the Center channel to anchor it on the screen, but sometimes i'll place a tiny percentage of dialogue L and R as well as using the center for the anchor. I do this sometimes with indie films that have very bad dialogue, so that the "noise" behind the dialogue doesn't just seem to be a SPOT sound, When I can completely fix the dialogue I'll drop it center and only move it around as the mix requires, but always keeping it at least somewhat anchored in the center. I'll also drop a bit of ambience in the center to smooth out the mix. I always remember sitting on the far left or right of the theater and feeling material that is just L and/or R not always feeling right due to the seat position, so a bit of Center can help that.
I've never run into panning L-C-R issues when down mixed, but i'm very careful about phase and downmix issues, always checking as I do sound design and mix. Some sounds I'll pan L-R some straight through the center. I play a lot with percentage of audio being sent to center while I pan L and R as well.

just do what you want and what the client/picture wants... but again. keep an eye on things and always test your down mixes.

cheers
geo

[iguana]

Hey, thanks again everyone for the insights. This is a goldmine for me.

Originally Posted by georgia

sometimes I've been know to build dynamic ambiances using various stereo sounds across all sorts of pairs, L-R, L-Ls, R-Rs, L-Rs, R-Ls, L-C-R-Ls-Rs, then I might very slowly move the various ambinces to build a bit of "movement" in a shot...

Actually, this is another thing I've been wondering about. I've recorded and built faux-surround ambiences in the past by putting different stereo recordings from the same location on the front and the back. This sort of works, but most ambiences that I hear in commercial movies sound much more immersive to me. Are these usually native 4- or 5-channel recordings, or is there another trick (that doesn't involve a $10,000 magic unwrapping box)?

When I can completely fix the dialogue I'll drop it center and only move it around as the mix requires, but always keeping it at least somewhat anchored in the center.

So when the actor exits stage right while talking, do you usually track him with the dialogue? I've heard both methods (tracking characters and leaving the dialogue pinned to the Center no matter what) in films, and they both seem to be a bit irritating to me when not done exactly right.

[georgia]

I create ambience with either libraries or field recordings or both. Either way I layer stuff.... layers layers layers. lots of tracks. I use ambience libraries, and then I layer in individual sfx... I do a lot prior to mix in sound design. If budget and time permit I'll do 5.1 field recordings... most of the time it's 5.1 libraries or stereo libaries, mucked and adjusted, Eq'd and compressed, or whatever I need to do. Yeah... I confess... I use my $20,000 system 6000 box to unwrap, and for reverbs on occasion in the process.

All the stuff in libraries and recordings are just the flower and water... if you wanna make the cake.. there is still lots of work to do to the ingredients... that's why they call it SOUND DESIGN and MIXING and not SOUND IMPORT and DONE.

Watching an Intern struggle with looking up a sound for 2 hours, then sitting down , pulling 5 sounds from the same list and mucking with them, layering them, combining them, to create the sound or ambience they were looking for in about 2 minutes.
that "OH!" look in their eyes the first time they get it....

LOL... Never gets old...



as to "tracking" dialogue... yeah once in a while. but the basic rule is keep the dialogue center. unless there is a damn good reason not to.
cheers
geo

[charles maynes]

Originally Posted by georgia

I create ambience with either libraries or field recordings or both. Either way I layer stuff.... layers layers layers. lots of tracks. I use ambience libraries, and then I layer in individual sfx... I do a lot prior to mix in sound design. If budget and time permit I'll do 5.1 field recordings... most of the time it's 5.1 libraries or stereo libaries, mucked and adjusted, Eq'd and compressed, or whatever I need to do. Yeah... I confess... I use my $20,000 system 6000 box to unwrap, and for reverbs on occasion in the process.

All the stuff in libraries and recordings are just the flower and water... if you wanna make the cake.. there is still lots of work to do to the ingredients... that's why teu call it SOUND DESIGN and MIXING and not SOUND IMPORT and DONE.

Watching an Intern struggle with looking up a sound for 2 hours, then sitting down , pulling 5 sounds from the same list and mucking with them, layering them, combining them, to create the sound or ambience they were looking for in about 2 minutes.
that "OH!" look in their eyes the first time they get it....

LOL... Never gets old...



as to "tracking" dialogue... yeah once in a while. but the basic rule is keep the dialogue center. unless there is a damn good reason not to.
cheers
geo

Geert, this post (Georgia's) really needs to be stickied....

[Sonsey@mac.com]

Originally Posted by charles maynes

Geert, this post (Georgia's) really needs to be stickied....

+1 on that!

[bizzle]

Originally Posted by georgia

...they call it SOUND DESIGN and MIXING and not SOUND IMPORT and DONE.

thumbsup

&e

ps, danijel: is this the thread that you were referring to? What are the rules of the center channel?

[danijel]

Originally Posted by bizzle

ps, danijel: is this the thread that you were referring to? What are the rules of the center channel?

Yes, that's it. There was another, however, even more elaborate AFAIR.

[kister]

Originally Posted by iguana

- Generally, how do you make a stereo source fit into LCR? Say you have a stereo ambience that you'd like to place in the front. Do you just put it on LR and keep it out of the Center, or do you feed a mono sum into the Center as well?

Hi
Grab yourself a few favorite DVDs and play them in your theatre, muting the centre track. then muting the Left and Right.
you should quckly get a feel for track allocation.
Generally foleys will match the pan information of the Dialogues.
If you were to send them left and right you'll not glue them to the screen action and create uncomfortable psyco-acoustic imaging.
If you have a heavy atmosphere behind a dialogue track. any cover for this track will have to be panned to the centre too otherwise you will get a stereo pumping effect. Personally (getting back to the question) I like keeping my atmospheres across the Left right LS and RS channels. But always check the dolby decode!

[DynaForte]

I should add...

I have experienced movies that use LCR panning instead of LR panning.

So sounds often start in C and go L or R as an object goes off screen.

I should also add that the average movie that is dialogue-based, for instance a comedy, is AMAZINGLY mono!

Seriously. If you watch a dialogue based film and mute the C, and then also try soloing the C, you'll realize that 95% of the sound is in the Center channel -- sometimes.



The likely reason LCR panning is used is because it helps stop a phasing experience, or biased experience for those not sitting dead center in the theater.


The Center channel seems to be used for everything happening DIRECTLY on screen.

It also often contains music at perhaps 6db or so less than the LR.

[Fhl]

Great post. Thanks for sharing!