the SalMar Construction, front panel, right

In an earlier post to this blog, I described in general terms the SalMar and the project before me. 

Now I’m getting down to business and actually learning how to play this instrument. I have an instructor:  Greg Danner, who is charged by the Center for American Music with restoring and maintaining the SalMar. In the late sixties/early seventies, Greg was a graduate student (in Engineering?) at the University of Illinois. As it turned out, he was working in a space adjacent to the space in which Sal Martirano was creating his instrument. Somehow, the magnetism of Sal’s personality began to pull on Greg, and he soon found himself immersed in the SalMar’s creation.

According to Greg, Sal was influenced by a book called “Digital Electronics for Scientists.” (See above clipping, which describes the book.) According to Greg, there is, in the Martirano archive collection housed by the CAM, a coverless and battered copy of this book – Sal’s personal copy.

Greg relates how Sal, who was a brilliant creative individual but not necessarily an electronics engineer, knew just enough to have attained the level of the engineering school sophomore. Clearly, he was aided by experts like Greg Danner in the construction phase of the SalMar project and beyond.

I can personally relate to this sort of enterprise and collaboration. The making of music, well downstream of pioneers like Sal, is very much wrapped up in topics quite far afield of the traditional music education. We might find ourselves called upon to learn electronics in both its analog and digital forms, computer programming, systems engineering, not to mention carpentry, pipe fitting, and glass blowing, all in the service of our art. The gifted graduate student is often an essential source of information and inspiration. So it has been a great privilege to have the opportunity to work with the SalMar in the company of one of its architects, and to hear reminiscences of Sal along the way. 

The working method with Greg is to mount a camera over the instrument and let Greg play and talk. I then follow his actions in slow motion and freeze frame later, listening to his descriptions and explanations repeatedly. Next, I try the concepts out on the SalMar when my schedule and that of the CAM permits. The SalMar, of course, is much too large, heavy, fragile and encumbered with its patch wires for any sort of casual moving. It will be moved to the theater just ahead of tech week. Work with the choreographer in the studio will happen via teleconference hook-up.

Here’s some of what I’ve learned:

The SalMar is operated by first of all crouching down beneath the control panel and turning on the power. It runs on ordinary line voltage. It has several power supplies, some of which are cooled by a fan. The fan is fairly quiet. The voltages generated are +5 volts for the logic circuits and +15 volts for the amps. While you’re down there, turn on the two integrated audio amps. In the CAM installation, the SalMar is heard through four ordinary hi-fi (?) loudspeakers. In the theater, we’ll gussy it up a bit. Sal used to play it through four Altec VOTTs. It also featured a matrix of 24 bubble-backed loudspeakers, of automotive style, which were deployed around the space. The matrix fed these speakers in a dynamic way, and indeed the four channel output is always in motion, panning randomly around in “circles.” 

Randomness is a theme of the SalMar. It has “a mind of its own,” a mind driven by all of those logic circuits, not all of which seem to be directly and immediately user controllable. The interactivity of the SalMar is, by philosophic design, only partial. The user may make suggestions. The SalMar may take them or leave them.

Seated or standing before the control panel, you’ll notice that quite a few of the lights have lit up on power up. Is this a feature or a bug? Greg? Step one, for me is to ‘clear the board.’ The machine is “played” by holding a grounding rail with one hand and touching the controls with the fingers of the other. The silver button below the lamp is the one one touches to activate that function or send that type of command or data. Touching the metal ring around the lamp cancels that action and turns off the indicator lamp. Sometimes, due to the “mind of its own” feature of the SalMar, all that can be really certain has happened when a lamp has been turned off is that, well, a lamp has been turned off.

Basic concepts:

The instrument features clocks, oscillators, feedback, and modulation. The clocks, of which there are four, are at the heart of the beast. No pulse, no sound. At the moment, clock I is on Greg’s bench awaiting repairs. The SalMar does quite a lot of damage on ‘only’ three clocks. Clocks are designated on the panel by Roman numeral. Oscillators are paired to clocks: I – 1,2; II, 3,4; III – 5,6; IV-7,8. 

There are other markings beneath certain functions on the panel. “MSB” and “LSB” refer to ‘most significant bit’ and ‘least significant bit’ respectively. This is digital logic lingo. Most is more of whatever, and Least is less. But sometimes, less is more. An example: when referring to clock speed, lighting up all of the most significant bit lamps results in a fast clock. But when routing a clock pulse to a feedback operation, the LSBs are important seemingly in inverse proportion. The activation of all four of the LSB lamps (there are only three MSB lamps – the machine has a seven bit architecture?) results in too much feedback. Oh, really? You can have too much of that good stuff? (Depends on what property you are feeding back.)

Well…this must be continued…