Reverb in Surround

Question

I was watching the Tron Legacy video on SoundWorks Collection and it brought up some questions for me and my wife in regards to surround mixing. Having never mixed 5.1/7.1, I did not think about how reverb reacts with the other channels, especially in 7.1. How does the reverb decay on a sound travel through the channels? Is that something that is automatic with a surround reverb plugin, or do you have to treat each channel differently? Do you even want it in the other channels, or is it for the front channels only?

In the SoundWorks video, they mentioned that to give spaces more character, they'd put different decays, slaps, etc on different channels. Subtle, I'm sure, but I like that attention to detail. Tim Prebble mentioned that as a bullet point in a previous surround question, and I was wondering if someone could speak to what drives the decision to use that method instead of doing something simpler.

Thanks so much!

Answer 1

Most film mixers use a variety of reverbs/delays depending on the scene/environment. Sometimes mono up the center or panned, stereo in the LR, or 5.0 etc.. Sometimes two stereo pairs with one set to a different pre-delay in the rear. There are a variety of choices to be made, and great mixers know how to create the right immersive sound with a large sweet spot, without muddying up the theater in a wet mess. With 7.1, they are given two more channels on the side to play with.

Say your scene takes place in a dark alley. A mixer might call up a general 2.0 outdoor reverb for the dialog in the LR, and a 5.1 reverb for the fx with an early reflection to simulate a narrow stone walled space, and add a mono slap delay panned to the left on some of the louder fx to simulate the slap off of a building. Lots of creative choices being made depending on the environment/storytelling.

On the typical dub stage, you'll see a single or multiple TC System 6000s, Lexicon 960s or 480s on the desk; with units strictly for the dialog/music mixer and units for the fx mixer. On a TC 6000 for example, you can set it up to run 4 stereo verbs or one 5.1 verb. You'll often see the mixers switch between the two through the course of a reel.

Answer 2

I'll be perfectly honest, I've never considered the idea of using multiple reverbs across individual channels in a surround mix. Now that I've heard about it though, it makes perfect sense to me.

Think of it this way, when you instantiate a reverb, there's little to no manual control over where exactly the reverb tail appears. The algorithm looks at the panned/positioned source and calculates what the decay is like across the entire space. So, it defines how it behaves in all of the other channels.

Now imagine how that works in a surround environment, perhaps you want to more closely emulate a real world, highly irregular, space...something that doesn't provide uniform propagation. Say you've positioned something in the rear and want it to propagate only in the rears and the sides; that there is some structural piece in the shot that would, in real life, cancel out or prevent propagation in the front (or maybe it just exhibits a different behavior to begin with). Working with only a single surround reverb would not afford you that sort of control; unless you had some extra return controls where the reverb is feeding back to multiple auxes.

Since adding more channels to mix isn't always the most efficient method, using multiple reverbs would gain you a simpler means to obtain that flexibility. It allows you to build a more complex and unique space. The trade off is that you need more processing power to do so.

Answer 3

This was chatted about recently on gearsluts post forum and one of the guys suggested running a 51 aux and inserting 6 multi-mono verbs across it. This is CPU hungry!! - BUT it gives you a panning 51 reverb, in that as you pan to that send, different proportions are sent through the individual channels. You just link the controls across all of them (default). Or you can tailor each channel as desired by unlinking. If you then widen the divergence on the source panning to make it slightly forgiving you get a verb that moves the decay around with the panning of the source, which isn't easy to come by in a DAW. The system 6000 does it, but that's a pretty penny...Revibe, TL space, altiverb (AFIAK) and most goto verbs apply a tail within the surround space and dont follow any panning of the source. Keep in mind, when you're editing stuff like this, frequency content and levels generally translate ok from a well designed edit suite to a dub stage, but panning is often problematic, so I wouldnt render this stuff without talking to the mixer or running some tests.

Answer 4

great thread.

FYI for any game guys the reverb in FMOD mixes the input to mono and returns on all 6 channels = lame

Answer 5

The most common reason to have lots of reverb units in a film mix is that film sound is mixed in stems and each stem needs at least one reverb unit. It's common to have separate stems for DIAL, ADR, Loop Group, BGs, FX, DESIGN, FOLEY, MUSIC, etc., as needed for the particular film and depending on capacity of the bussing and the recorder(s). So in this hypothetical stem count of eight, you'd need eight reverb units (or separate engines) at minimum to keep the reverbs segregated by stem.