By 1991 automated lighting boards were common in the theater. A lighting board turned on specific lights out of hundreds, each to a specific amount. Each lighting change took multiple dimmers being manually raised or lowered to a specific amount. Now with lighting automation, the changes to the dimmers were all changed by a computer. Press one button and the computer would fade up and down any number of lights all at the same time, following the preset set of instructions written into the lighting board’s program.
But at that time there was no such comparable automation in sound. Prior to any sound automation, every microphone had to be manually raised and lowered. Oftentimes a few of them were raise and lowered on a line by line basis. Few sound mixers could accurately pull it off. And with the shows’ sound getting more and more complex, more and more inputs had to be raised and lowered on a note by note, line by line basis.
Mixes done in the studio were getting more and more complex as well. But automation had already existed for the studio. The difference between in the studio and live was simple. In the studio once the music or sounds were recorded, they all played exactly together from a tape, or hard drive, repeatedly. The tracks or inputs in the studio were, in the same place relative to each other, and at the same level, every time. So applying an automation system was easy. The tape or hard drive, output a code of time, “Time-code”, and the changing levels of the inputs were controlled by a computer “locked” as we say, to that time. Piece of cake! Very repeatable! The automation followed the same time-code every time. We called this “in sync”. (in Synchronization with the taped performance).
But when it came to a live show, there was no Time Code. The show sped up and slowed down, slightly but enough to make problems. The songs would change their speed imperceptibly to the ear but enough to fall “out of sync”. Each live actor or player would also change their performance, every time. No time code could run along with a live performance, which could control the computer that would adjust the levels. No time code could make changes for an actor being hoarse or substituted, or a musician being substituted. A live performance always speed up and slows down a little, not so much that the audience hears it, but too much for a computer to control it. We called this “out of sync” or un-synchronized.
So I had an idea. Some things are inherently timed by nature. Like when a dancer leaps into the air there’s a good likelihood they are coming down to earth in the usual amount of time. Other things take consistent amounts of time too. Like a song might be 100 bars or more in length, and vary by seconds, but every 8 bars of that same song doesn’t vary by much. So I figured if a mixer/ operator re-synchronized a shows computer every few bars of music, or on the starts of some songs, or a key places or scenes in the show, that show’s computer could be mostly in sync with the show. And if there were indicators of changes in overall levels from substitutions, at the start of a show now the automation could handle those changes as well. Now a show’s sound automation computer could re-synchronized by the operator/mixer very often, and therefore be (pretty much) in sync with the show.
I had to test my theory. I took a couple of time code generators from the shop and fed them into a time code reader. I set the code reader to chase. This meant the reader would adjust to the incoming time code. Then I kept switching and advancing the incoming code by advancing the time-code generator. Sort of like one code device following another time-code device rather than following continuous code generated by a tape machine. It worked like a champ. It may seem obvious now thirty years later but then, no one had thought of it. Everyone said a live performance varies too much and you cannot chase a live performance. They were wrong.
I contacted Andy Smith an automation designer for music and film studios and asked if he could make a few changes to his studio/film mixing program so the code reader would follow itself, updated by the operator. He said no problem, and we together invented “Performance Automation” the first automation system to follow a live performance.
Now we had to use this contraption in a real show. I had been hired to design what was the biggest musical on Broadway that year. It was called City of Angels. One of the same producers that had co-produced Sophisticated Ladies, (the show I saved), was the producer of City of Angels.
Sound Associates financed the purchase of a new console for this show, which hopefully it would be ready in time for this show. It wasn’t. City of Angels was supposed to have the first Theatrical Automation but it wasn’t ready.
We then put all the parts into a stand alone box that could be looped through to any console, and make that console “automated”. These parts consisted of 24 motorized faders that controlled the fader’s position by a digital controller, but all the audio followed the analog path also through the fader. And in order to make this work over the distances that were between the external fader and the faders within the existing consoles, a special analog amplifier had to be added.
Well the first show we put this into was Will Rogers Follies. I could sit in the middle of house (theater) and balance out all the orchestra on a bar by bar basis. Then the faders, and computers could be moved back to the mix position, and then have the automation repeat that mix, in sync, night after night.
It all worked perfectly. Interestingly, this first time we use used this system, we had 300 feet of audio cable between the remote fader package and the actual house console. When we removed this cabling to put this remote package near the house console at the mix position, the sound was much “brighter”. Obviously the cabling had rolled off some top end. We ended up adding back the several lengths of cabling and leaving it in the system for the entire run of the show.
There system ran flawlessly for the entire run of Will Rogers Follies. Ned Hatton was the Mixer for the whole run.
This Automation system had some interesting additional features. If you touched a fader during any cue that fader would not “fight you” (try to move), It would simply drop from the program at that moment, just for that cue, and allow the operator to adjust it manually himself. This way if there was a substitute musician in the Orchestra Pit who didn’t know how far to sit from the mic or perhaps wasn’t as familiar with the score, the operator could adjust for it.
Also the system sent out midi and time code pulses to trigger every kind of external device. Now the go command on the automation system operated everything.
The clock inside the automation system was adjustable in “sub-frames” which means that every change would as close as 1/3000th of second part. Fast enough to do anything.
The fader speed was also remarkable. Every fader could move at the rate of 2000 dB ber second. This is really fast. Fast enough to take a pencil, place on a fader with the fade completely down and program the fader to fly up as fast a possible. With this you could launch however many pencils you could load, and fire them across the theater, into the theater seating, during tech. (but don’t get caught)
One incident happened during this first run. The cleaning service had come into the theater and apparently knocked the computer off the shelf. They simply put it back on the shelf and didn’t tell anyone. They had done no permanent damage but knocked the timing board out of its computer slot. During intermission one day I got a call from Ned. He said “the system is not working properly, the timing is all off”. My immediate reaction was shock. I asked Ned what did the diagnostics say when you booted up the system. Ned said “it said it wasn’t working” I said then why didn’t you just not use it? Ned said “It always works, so I didn’t believe it” And that was the sum total of all the failures the system had in all the shows it had done.
Bernard Fox
