We have 4 speakers arranged as follows:
The two front speakers are Meyer UPL-1s, the speakers that normally hang at the front of CNMAT's main space. The rear speakers are Meyer HD-1s, the speakers that are normally in the rear studio downstairs, resting on a pair of our large tripod speaker stands.
We use a Mackie CR1604-VLZ mixing board to mix sound for the concert. (Photo below.)
The outputs to the speakers are from the 4 sub outputs of the board, so the signal in sub 1 goes to speaker 1, etc. (What Mackie calles "sub" mixes are also known as "bus" or "group" mixes.) Using sub outputs for individual speakers somewhat limits the ability to dose each sound source independently to the four speakers; click here for details.
The main (L/R) outputs from the mixing board go to a DAT machine (Tascam DA-P1) so we can make a tape of the performance.
Aux sends 1 and 2 from the mixing board go to the reverb unit, a Lexicon PCM90.
We use all 16 inputs to the mixing board. All inputs are balanced, and unless otherwise noted all inputs use the 1/4 inch line inputs.
For dynamic sound spatialization, we use hardware and software from Level Control Systems (LCS), with additional custom software by Matt Wright.
The 4 outputs from the LCS system go to the 4 speakers in the hall respectively, via the stage box at the front of the stage, the downstairs patch bay, the stage box at the rear of the hall, and the mixing board. (Signals to the front speakers then go back via the stage boxes and downstairs patch bay to the front of the hall.)
The LCS hardware has 8 inputs, of which we use 6 for the concert.
Campion's 6 sources to be spatialized are signals from SampleCell cards in his Macintosh.
Zbyszynski's 3 sources are:
In order to change what signals are routed to the LCS inputs without having to rewire in the middle of the concert, we use a patch bay. The bottom half of channels 1-6 are wired to the LCS inputs. The top half of channels 1-6 are wired to Campion's SampleCell cards. Thus, when there are no connections on the front of the patch bay, Campion's synth signals are "normalled" to the correct LCS inputs.
Zbyszynski's sources are also connected to the back of the patch bay, in the top half of channels 7, 9, and 11. No connection is made to the top half of channels 8, 10, or 12. Thus, when patch connections are made from the top of channels 7-12 to the bottom of channels 1-6 (respectively), Zbyszynski's sources are connected to the LCS inputs.
Channels 8, 10, and 12 are "silent" sources, which we connect to channels 2, 4, and 6 to break the "normal" connection from Campion's synths during Zbyszynski's piece. This is so the noise from the SampleCell cards (which is substantial) won't be heard during Zbyszynski's piece. (And also so that in case Campion's synths suddenly start making a sound they won't be heard.)
So at the start of the concert, for Zbyszynski's piece, there are 6 patch connections on the front of the patch bay. To reconfigure for Campion's piece (in the middle of the concert), all that we need to do is remove the patch connections from the front of the patch bay. (We could have done it the other way around, but we felt that removing connections was easier and less error-prone than making connections.)
Here's how the rack looks when configured for Campion's piece:
(Some other interesting details shown in this picture: two 8-channel audio snakes come into this rack, one from Campion's SampleCell cards, and one from Zbyszynski's synths. Since we only need 6 channels of the 8-channel snake, we tape back the extra two plugs; you can see this on the left of the picture, with the orange and green plugs. There is a ProCo direct box immediately above and behind the patch bay, used for the TG77 output, which is sent directly to the mixing board rather than through the LCS. Hanging in the air to the left of the rack are the outputs from the CD player, which were plugged into the front of the patch bay so that Bill Frisell music could rotate around the audience before the concert. Behind and above the direct box you can see the wall power outlet that the entire system is plugged into. )
Here's how the patch bay looks when configured for Zbyszynski's piece:
LCS Cue Control software runs on a Macintosh and connects via MIDI to the rack-mount hardware. (We ran version 2.46 b4 for the concert.) The data transferred on this MIDI connection is encoded in a special custom format.
We take advantage of LCS's abstraction of "SpaceMaps." We create a SpaceMap by describing the location of our speakers to the software. This defines a 2 dimensional space representing the room as viewed from above. LCS can then place any sound source at any given point in this 2 dimensional space, by controlling the relative level of that sound source in the appropriate speakers.
LCS has a MIDI implementation where the first 24 MIDI controllers map to the X position, Y position, and overall volume of the 8 sound sources to be spatialized.
We connected Max to LCS Cue Control on the same Macintosh via OMS's "IAC" bus.
The custom Max patches output continuously-changing X and Y coordinates for the 6 sound sources used in the concert. The inputs to the custom Max patch are 8 continuous variables affecting the parameters of the spatialization processes, which come from a J. L. Cooper Fadermaster operated from the middle of the audience. (See photo below.)
Click here for more info on what the custom Max patches actually do.
Everything MIDI is at the front of the hall: WX7, Disklavier, The Mac for Campion's and Zbyszynski's Max patches and Campion's SampleCell Cards, Zbyszynski's synths, LCS hardware, Macintosh for LCS. This results in a big pile of gear at the front of the hall, but makes it reasonably easy to wire everything.
This table of equipment sat on the stage, to the right of and behind the piano. This is the Macintosh used for the two interactive pieces; the monitor sat on top of the rack so it would be visible to Zbyszynski over the top of the piano. The rack contains synths for Zbyszynski's piece, the Studio 4 MIDI interface for both pieces, and a patch bay we didn't use. Note the cardboard sticking out of the front of the rack between the TG77 and the TX81Z; that was our kludge custom keyboard stand. Not visible behind the rack are a CD player (for pre-concert music) and a mixing board (for Campion's synths.)
Campion used a Yamaha DX100 mini-keyboard as an additional MIDI controller for his piece, to trigger behaviors of his Max patch:
Matt's real-time spatialization control interface, a Fadermaster, is in his seat (on his lap) in the middle of the concert hall, connected by a long MIDI cable to the LCS computer.
Everything operated by the "sound man" (Ron Smith) is at the back of the hall: Mixing board, ADAT, Reverb, DAT.
Tom Swafford played violin for his piece "Sight Unseen." He started the piece in the rear studio, where there was a microphone wired to the mixing board and sound system upstairs. So at the start of the piece the sound of his violin was heard only through the speakers.
Then, as the piece went on, he walked, still playing the violin, to the outside door at the end of the hallway, which was open. The windows to the main room were also open, so the sound could be heard coming in from outside. Tom then walked down the hallway on the bottom floor to the other outside door, and out into the yard below the bridge. Then he came inside and walked up the stairs to the landing. Tom went back to the microphone to end the piece, never having shown his face in the concert hall.
Connections from the rear of the hall to the front of the hall:
Connections from the front of the hall to the rear of the hall:
Connections from rear studio to the rear of the hall
Here's a picture of the stage box near the mixing board at the rear of the hall. At the bottom of the frame you can see cables sticking out from under the rug.
To avoid ground loop problems, we connected all power to a single source. We used one of the wall outlets behind the stage, choosing an outlet marked with a black "X". (This is visible behind the picture of the patch bay above.) All the equipment on stage was connected to powe strips connected to this outlet. A long extension cable brought power from a power strip connected to this outlet to the rear of the hall to power the rear speakers, mixing board, ADAT, etc.
There were 3 pieces on the program that used the ADAT: Smith's (played twice on the program), Pitter's, and Campion's. Each of these pieces uses a different number of ADAT channels (4, 6, and 5, respectively), and a different way of routing the ADAT outputs to the 4 speakers. This caused a lot of work for the sound man to switch between configurations, including a lot of somewhat annoying audible button clicks during the concert. Next time we should institute some convention about multi-channel tape and have all the composers stick to it. E.g.,
Each piece that used the ADAT had its own tape, so there was additional delay starting those pieces while new tapes were put into the machine. Perhaps in the future it would be better to make a single ADAT tape with the music for all of the pieces.
There was a disturbing amount of delay between the time you changed a spatialization parameter on the FaderMaster and when the effect of that change became apparent.