Through most of his analog works from Sidewinder on, Mort Subotnick relied heavily on the use of analog control tracks to pace, locate and otherwise manipulate sound events through time.

There have been several explanations given by many people for his use of this technique, most predominant among them being they afforded a level of control, predictability and repeatability which could not be had otherwise with the technology avaliable at the time. There is also the logical conclusion that their use in effect increased the size of the instrument he had to work with, freeing components for sound processing that would otherwise be used for controlling the parameters now committed to tape. The idea behind the process is simple: an audio track is recorded onto magnetic tape and then fed through an envelope follower, where it's amplitude is converted into triggers and control voltages. Thus a 'ghost score' is created which in turn is used to 'play' the synthesizer at a later time - much like a player piano or to a certain extent, midi.

One can imagine both the benefits from such an arrangement and the difficulties they introduced. Working with a series of recorded sine wave bleeps requires a thorough understanding of the finished piece as these precompositional stages do not even closely resemble the net result. Mort was indeed a master at this, using these tracks to determine not only the start of an event, but it's spacial location, amplitude and frequency components.

His use of control tracks became increasingly more sophisticated and in time he would dedicate four of the eight tracks of a multi-track tape machine to these control events... and at times this wasn't even enough. Addressing this, he developed a process by which multiple control tracks could be recorded onto a single magnetic track of tape. Using the Buchla comb filter (aka fixed filter bank), he would record at times up to three independent tones on each track. Pure tones all with a distinct frequencies, they would be demultiplexed at playback though the filter bank, thus creating three independent and isolated control tracks from each track of tape. Keep in mind the Buchla comb filter is unique in that is had separate non-attenuated outputs for each filter, allowing for this type of decoding. This process led to Don Buchla's development of the 296 Programmable Spectral Processor, a comb filter in which each of its 16 banks each had it's own envelope follower.

For my buck, Subotnick's most explorative use of control tracks was evidenced on Until Spring and Sky of Cloudless Sulfur (SOCS) and I will long remember the first time he played sections of SOCS for his students at Cal Arts. He started the tape machine, and within a second or two of the 'dance' section, our collective mouths were on the floor. Many voices - obediently synced to one another in space and time, all moving in unison, all parts of a much larger whole. Unforgettable, and at the time, quite unbelievable. Magic was happening. I remember Jill Fraser looking over to me with almost a look of panic on her face. Click here to hear an excerpt from the dance section of Sky of Cloudless sulfur, part 2. (Used by permission of the composer - see below)

We knew the size of his instrument, we knew how much sound could be had in a single pass, yet we were hearing what appeared to be a quartet of instruments playing along with one another in perfect unison when he wanted, and sharp rhythmic syncopation when he preferred - none of which could have produced in real time. Keep in mind this was six years before the development of MIDI. If nothing else, it was all flying along faster than any human could play and as big as his instrument was, it wasn't THAT big. Yet in reality, through control tracks he had indeed made it that big.

And so it was in SOCS, realized on an eight track Ampex machine, with four of those used for control tracks in command of the remaining four tracks we heard.

His methodology was as interesting as the final result. Sometimes he would simply record metric pulses to be used later for triggers to pace and sync sound events. There were many of these, each picking off different divisions of musical time (halfs, quarters, eighths, sixteenths and triplets). At other times he would just sing, grunt, 'r-r-r-r-r' or otherwise hurl into a mic and record that information, the variations in amplitude being then converted into tracking control voltages by the Buchla 230 Envelope Follower.

Among the most creative gestures were a series of of pleasant, yet vaguely comical short flutter-tounged phrases used throughout the piece. Part two begins with one of these little guys and you'll know exactly which phrases I am speaking of when you hear it - rising as if posing a question before the beginning of the dance. He literally flutter-tongued (like rolling one's 'r's") to produce these elements. Rises in amplitude were used to control the pitch and pan and the 'tuhs' themselves converted to triggers for the notes within the phrases.

Another hallmark of this work are the 'chiff' transients in the percussive events and a while back I attempted to duplicate this as precisely as I could on my analog system. Fitted with Low Pass Gates I was fairly sure I was armed with the secret ingredient (more on this in a second). I assumed these percussive chirps were done by simply applying a short transient via an envelope generator to 'twap' the frequency of the sound as it began. I soon realized however that while this worked well with slower paced repetitive lines or single beats, it become very noisy as their period increased (note value decreased). I then attempted the same transient on the filter, and while that was much closer, it still wasn't it. Finally, I mixed in a touch of white noise with the main oscillators to augment the filter transient. BINGO! I was convinced I had nailed it. Regardless of the frequency of the sound or it's period of repetition, they manifested as naturally as Mort had carved them some 20 years back.

Confident with this discovery, I dropped him a note to see if I was right. His reply came shortly afterward and was not what I wanted to hear, but educating none the less:

"I used a mixture of oscillators...but I did not use any noise. What I did was put an envelope follower with a transient up and v quick, almost transient down mixed with another voltage which gave it its pitch...I had voltage mixers [I think they became standard]. That is what created the transient at the beginning of the sound...it was a transient!"

Those damned control tracks once again! It seems that the secret was four fold: 1) the response of the Buchla Envelope Followers and their ability to track relatively hurried attacks efficiently, 2) The voltage curves generated by the Buchla envelope generators, 3) the response of the freq VC inputs of the 258 and 259 VCOs and 3) those low pass gates. And while in theory you would think that Mort's use of control tracks would allow SOSC to be reproduced on virtually any synthesizer, in actuality it was all about the Buchla's attributes, strengths and weaknesses. Indeed, I will go on a limb here - focusing on technical issues alone and for the moment ignoring its compositional prowess, (which to my aesthetic runs quite deep), SOCS would be impossible on any other instrument on four counts:

The 230 Triple Envelope Follower, the 258 and 259 VCOs, model 292 Low Pass Gate and to a lesser extend, the 280/281 Quad Function/Envelope Generators. While the Envelope Follower I think has been sufficiently discussed, the Low Pass Gates, VCOs and EGs are equally suspect in giving SOCS its unique bouquet, to wit:

A topic of much heated discussion within the small circle of synth DIY'ers, the secret of the 292 Low Pass Gates are the vactrols. A small opto-electric device used throughout the Buchla 200 which some will rightfully say have no place in a VCA design, which in theory needs to be fast to be efficient. Yet due to the vactrol's naturally slow reaction time (most notably an inherent 35ms decay), they afford a unique ringing of the low frequency spectra which provides the acoustic personality of the percussive phrases used in SOCS. The net effect was quite encompassing and extraordinarily natural. As reference, Don Buchla wasn't the only instrument designer experimenting with these unique devices, as vactrols can also be credited for the warmth of the Moog Phase Shifter.

The linearity of the 280/281 EGs also played a part in the 'thwap' transient.. It's the linear curves. Attempting this effect on another manufacturer's offerings, especially most ADSR's will not yield the same results because of their logarithmic response. A Blacet EG1 will come close, but one will need to VC the attack (while it's set to zero) with its own output to get anywhere close.

Lastly, we must not overlook the VC freq response of the Buchla VCOs as a contributor - the secret here being their triangle wave core -and- the architecture of the OTA's driving their expediential converters. To attempt this with any other VCO will be futile without a hell of a lot of external processing of the transient before it makes it's way to the VCO.

In closing, while Mort Subotnick must be wholeheartedly credited with the originality and uniqueness of both Until Spring and Cloudless Sulfur, Don Buchla had a bit to do with it, too.