|home||Space Station SST-206||consulting and articles||SST-282 History||Ursa Major Documents|
Left to right: Chris Moore (Ursa Major), Tony Agnello (Eventide),
Wolfgang Schwarz (Quantec), Dave Griesinger (Lexicon)
Left to right: Dave Griesinger (Lexicon), unknown person,
Chris Moore (rear row) (Ursa Major), Barry Blesser (EMT),
? Sekiguchi (Sony), Bob Orban (Orban),
Space Station demonstration cassettes
Three demonstration cassettes were made for the Space Station in the period 1979 - 1981.
The genesis of the Space Station
In the Spring of 1977, I resigned from my position as Project Engineer at Lexicon. I set out with several goals in mind. One was to design the first low cost digital reverberator, another was to work in my own company, and the last was to establish the company framework in such a way as to be fair and motivating for each of its workers.
By the Fall, I'd read a book on digital hardware design (up until this time I'd been an analog guy), cleaned up my cellar, and established an office and several lab benches. I began to sketch out a hardware design for the algorithm I had in mind, which was a single delay line with one write and eight reads. A key idea of the Space Station was that if the delay times of the taps fed back to the input were slowly varying as a function of time, the transfer function of output to input would be constantly changing and would tame the peaks that otherwise (in a time-invariant system) would have occured, dooming the algorithm to oscillation, useless for reverb.
When the hardware was up and running, the first listen was very discouraging. Stability was enhanced, but there still was ringing and, worse yet, a gritty sound as the taps wandered in time, skipping 62uS at a time. Digging a bit deeper, I found that, through my error in marking the punch card that programmed a small PROM to control the modulation, one of the eight taps was not modulated. Listening only to that one stationary tap, the sound of the reverberation was much cleaner--I saw and heard that if some of the taps were modulated and confined to feedback duty while others were confined to "auditioning" the interior of the "room" and not modulated, the grittiness and other anomalies would be reduced.
Now I entered a redesign phase where I stretched the envelope by requiring fifteen modulated feedback taps, eight unmodulated audition delay taps, and one tap dedicated to special effects (Echo mode). The ADC/DAC job was suddenly tougher: in 62uS (one sample time), I had to sample and hold and convert an input from analog to digital and convert from digital to analog via sample and holds 24 tap outputs. Furthermore, the converter needed to have a wide dynamic range (I was shooting for about 14 bits worth, or -84dBFS), highly precise gain input to output, and good linearity. I called into play several techniques. One was to use an 11 bit instantaneously companded approach (sign bit, two range bits, and 8 mantissa bits), another was to time multiplex all the conversions, first doing the ADC conversion using the DAC as the core DAC in the successive approximation ADC, then using the core DAC to do 24 DAC conversions for the taps. This trick ensures reduced distortion and precise gain management and could be implemented using only an SAR and two inexpensive 8 bit DACs. The results were good and proved to be stable in manufacturing. I could go into detail about a number of other key design problems and issues, but this is enough to give a taste for what was required for the design.
Booth 100B at first AES Convention, May 1978 Chris Moore at first AES Convention, May 1978
By Christmas 1977, the
prototype was running and I was tuning equalizer curves, delay values,
modulation range and rates, etc. The prototype was built on Vector boards
that plugged into a card file with a wire wrapped backplane. Next, I packaged
the card file in sheet metal to fit a 3U rack mount enclosure with user
interface controls and a professionally painted and silk screened front
In May 1978, I took the single prototype and some simple display elements and flew to Los Angeles to exhibit at the AES Convention in the LA Hilton. The show was encouraging, but not dramatically so. One trend which emerged as early as this was that distributors and studio engineers from overseas were more interested than their domestic equivalents.
Meanwhile, the prototype was being transformed into a PCB version that would end up having four PCBs and hundreds of ICs. In this phase, as indeed in the design phase, I was aided by two friends who had remained at Lexicon but who were interested in working in their own startup. They helped some with design, but mostly with logistics (getting sample parts, establishing credit with suppliers, building and wiring, using up some of their cash, etc.). Such was the care and accuracy with which we worked that the first batch of five PCB sets were all correct (only two etch cuts and jumpers!) as was the sheet metal. We were able to sell the "protoype" run as the first five production units. Our first shipment was to 3M France, in roughly November 1978. After the first five systems, we built Space Stations in batches of 25 and found ourselves able to begin payroll (finally something coming in!) in January 1979. We were solvent and well on the way in the marketplace. The Space Station was not changed during its entire product life (well, we did change an op amp to one with a higher slew rate) and went on in the next eight years to selling somewhat over 1900 units.
Excerpt from first Space Station Brochure
My reactions to our
success were varied. On the one hand, I was proud that we'd gotten off
to a good start, that the product was being well received, that we were
organized and working as a cooperative, and that our financials looked
good. On the other hand, I had not been able to tame the various flaws
I could hear in the Space Station. These included spectral smearing (due
to the wandering feedback taps), modulation noise (delay taps were simply
picked up and moved 62uS with no smoothing), and the inability to get a
really distant sound due to the fact that the Audition Delay taps by design
picked up the dry source as early reflections as well as the dense later
reverberation. I was concerned enough by these shortcomings that I stated
one of them in a caveat emptor in the first version of the brochure. I
also, by relatively simple means, broadened the capabilites of the Space
Station by allowing for a rich, adjustable set of special effects that
could be combined with the reverberation. In essence, I was shaping the
product so that a customer who might not find the reverb suitable in some
situations could still find many useful possibilities. Our English dealer
at the time dubbed the Space Station a "reverb effect" to distinguish it
from a more powerful reverb (but these were selling for upwards of $7,000
at the time, compared with the Space Station at $1995).