Ahh, surround in live venues can wonderful, for many things beyond video playback.
When doing musical
theatre sound reinforcement in the 90's, my default
FOH drive racks included a Miles Tri-Sonic MTI-3A Imager.
We used these units in venues ranging from 400 - 700 seats as an affordable way to generate Left-Center-Right-Surround (LCRS) from a stereo band feed, which was converted into LCRS, plus a
mono vocals feed that
fed center only, and a direct feed to surround, for
reverb and effects.
I've also heard great surround / distributed sound in smaller venues as well.
There are likely low-buck ways to generate the same signals using
current consoles, to use the normal stereo pan mixing approach to achieve LCRS. (Discrete Quad is another approach, but I found that less workable for typical stereo-based shows.)
This thread is about surround sound, but as a side note, Well done
LCR is terrific, providing an even and smooth sound
stage for stereo music for the entire audience, not just those sitting down the
center line, with firmly planted center vocals. This works really well for musical
theatre. After having used
LCR setups, either standard L-R only or
mono clusters only each fall far short, for the sound localization for those audience members not sitting at the center of each aisle. (Those Dolby guys who added multi-channel and surround to the 'talkies' way back when had it right.)
We used surround speakers to enhance the sound of the band and vocalists in the room, and also for sound effects.
The stereo band mix
fed difference signals (L - R) into surrounds (e.g. only stuff purely in left, or purely in right, nothing panned center), and vocal reverbs and ambient and special sound effects could also be routed only to the surrounds.
The actual source of the ambient sounds makes a big difference, and feeding
reverb into the room from the back/sides just into the surrounds allowed us to 'load' the room with more, great
reverb, without
feedback and without losing the vocal presence and intelligibility. This was a very minimal version of some of active acoustic enhancement systems.
Likewise, sound effects work very well (not unlike the movies) for both ambient and specific effects, as humans can easily localize the sources, and it really does envelop the listener. The effects can only be pushed so far, though, or they break the '4th wall' and pull people out of the show. (e.g. quiet thunder and rain all around draws you into the show, but big thunder and rain all around makes you think there is a storm outside and worry that you didn't close the car windows.) While it might seem that more directional effects from a single source would not work through a multi-speaker surround
array, even those effects worked well as being generally behind or to the side of the audience, even if different audience members heard the sound as coming from rear left, rear right, or rear center, depending on where they sat.
We typically ran a center
cluster that covered the
venue, along with left and right sides that covered most seats.
(We typically also ran
front fill,
under-balcony fill, balcony fill,
etc., for intelligibility in tough rooms, but that's secondary.)
Surround speakers were smaller, and worked better if they had a wider dispersion, with 4 or more speakers at a lower volume, rather than two larger surrounds. 6 - 8 sounded really sweet. They did not need to be high
power, and we sometimes connected them in series/parallel to reduce the amount of amp channels.
The MTI-3A from Mike Miles was a great
unit that seemed to combine concepts from the original stereo-based LCRS Dolby Optical Movie 4.0/4.1 Theater Surround sound, along with concepts from Bob Carver's Sonic Holography.
Here are links to the archive of the Miles information, including the
AES pre-print that includes the math:
http://www.milestech.com/
http://www.milestech.com/mti3.htm
http://www.milestech.com/tricompa.htm
http://www.milestech.com/aesimpre.htm
Early 4.0 stereo-optical surround sound added Left and Right to get Center (
Mono), and subtracted Right from Left to get Surround. If I recall, the Optical Surround processing of stereo-optical movie
theatre tracks that conveniently ended up on many movies on VHS format also included Dolby-B noise reduction, some delay to the surround channels, and band limited (e.g. included only 100Hz to 7KHz) the surround channels. The decoder was implemented in home receivers. This was prior to the implementation of pro-logic 'steering', and prior to stereo surround (5.1, 7.1) channels. (The whole interesting story about how we got the .1 LFE channels also starts back at optical movie soundtracks...)
(Side note - One early 'low buck' way to generate surround was to just hang a high
impedance speaker across the stereo
amplifier's L+ and R+ outputs, which only reproduced the L - R sounds. (When you bridge an amp, you
wire it the same way, but then invert the input on one
channel, so they add. This skips the invert, so they subtract.))
My take on Carver's sonic holography was that it was a way to make speaker-based listening sound more like
headphones, to get rid of the extra sound from the right
speaker heard in the left ear (and vice versa), by adding some 'negative right' to the left
speaker with appropriate triangulation delay, and vice versa. When I heard it, it sounded great for home hi-fi, but had a super small sweet spot (as in
bolt the listening chair down to the floor.)
So, as I understood it, the Miles
unit took a stereo input, and created the following:
Left Out - Left In with some negative Right In
Right Out - Right In with some negative Left In
Surround Out - Left In Minus Right In
Center Out - Left In Plus Right In
There was a matrix control that adjusted how much of the opposite side was included, and also how much L+R went to center.
If you turned the knob all the way down, you got plain Left into Left, plain Right into Right, and nothing from L and R into Center.
(The Milestech
AES pre-print has the equations and math as to why it all sums acoustically.)
This setup made stereo band mixes sound great, because you could get even coverage across
LCR, and the panning was just right. Even if you had something hard panned to one side, the 'negative' amount added in to the other side meant those on the other side of the
house still heard the
instrument. This solved the classic PA stereo issue of the audience on the sides not hearing a full band mix, because they were missing the sounds panned to the opposite side.
Then, the Miles
unit had a direct input for Center (where I
fed vocals from an per-channel assignable
mono bus (not a
console sum of L+R) or aux-fed
mono feed, if there wasn't a true assignable
mono bus.)
Finally, there was a direct input for surrounds, which I
fed from an aux bus that included input-channel returns for reverbs,
etc.
I generally found the setup to sound really good without a lot of fuss for each
instrument.
It would seem to me that one could re-create these signals easily in modern consoles. For surround, bring the L and R channels back into unused inputs, invert the right
channel polarity (or use a pin 2-3
polarity swap
XLR barrel), and only assign those two channels yo a
pre-fade aux bus, and out to the surround amps/speakers. Similarly, you could create the 'negative' L & R feed into the opposite side with inverted channels, but would need to introduce the inverted opposite
channel downstream of the L&R busses (e.g. into a matrix or similar) so you aren't introducing electronic
feedback.
This approach tends to require extra work for both installation of surround speakers, and by providing overlapping (rather than zoned)
LCR arrays. The two approaches can be used independently, and worked really well for us.
A final note on
LCR speaker sizes. I think arrays/speakers of the same size/family are preferred, but this can be done with a smaller center, with mostly vocals, along with a larger set of L-R sides. Likewise, can be done with a a larger center, where it handles more of the L+R load. The Miles 'Matrix' control also could be used to adjust that. In any case, the center does have an important
role, and needs a decent
speaker/
array.
-larry-