Flying Spots and Photomultipliers
The wave scanner in the Oramics machine is built from two components, and neither of them was invented for music. One makes a dot of light that can be put anywhere, very quickly. The other notices light arriving and turns it into a voltage, also very quickly. Wire the second to steer the first and you have an instrument.
The tube, used backwards
A cathode ray tube fires a beam of electrons at a phosphor-coated screen, and where the beam lands the phosphor glows. In a television the beam is swept across the screen in a fixed pattern and its brightness is varied to draw a picture. In a flying spot scanner nothing is being drawn: the tube is a lamp whose position you control. Two voltages, one for horizontal deflection and one for vertical, put the dot anywhere on the face of the tube, and they can move it as fast as the electronics driving them.
That is the useful property. Put something transparent in front of the tube and the moving dot illuminates one point of it at a time, which means a single sensor behind it can read the whole thing in sequence. No motor, no mirror, nothing to wear out, and fast enough to do it at audio rate.
The sensor behind it
The light that makes it through a painted glass slide is not much, so the thing reading it has to be sensitive, and it has to react fast enough to follow a waveform. A photomultiplier does both. A photon strikes the first plate inside the tube and knocks off an electron; that electron is accelerated into a second plate and knocks off several more; those are accelerated into a third, and so on. The RCA 931A used here has nine plates in the chain, and the multiplication compounds at every stage, so a trickle of light at the window comes out as a current you can do something with. The diagram below shows three electrons per plate rather than the real number, which would be unreadable.
The diagram
Five steps, with the arrow keys or the controls. The first lets you drag the spot around the face of the tube to see that two voltages place it anywhere. The third has a slider for the amount of light reaching the photomultiplier, which is worth taking to zero and back. The fourth closes the loop, and the amber line underneath is the sound that comes out of it.
A cathode ray tube used as a steerable light source, a photomultiplier reading what gets through, and the feedback loop that ties the two together.
Open the diagramIt needs a wider screen than this one.
Closing the loop
On their own these are two laboratory instruments. What makes them an oscillator is wiring the output of the photomultiplier back into the vertical deflection of the tube, so that how much light is getting through decides where the spot goes next.
Put an opaque shape in the beam's way and the loop settles at the edge of it: the spot is pushed upward while the paint blocks the light, and stops as soon as light gets through. Sweep the spot horizontally and that settling point traces the upper boundary of whatever has been painted, continuously, without anything ever having to work out what the shape is. The voltage doing the pushing is the signal.
The same trick appears twice in the machine. In the wave scanner it reads the painted glass slides that supply the timbre. In the film programmer, until 1972, small cathode ray tubes with the same arrangement followed the hand-drawn loudness lines on the control film. Those circuits needed a steady mains voltage, and when the supply at Tower Folly sagged far enough Oram gave up on them and fitted plain lamps and photocells instead, which is why the later graphs had to be filled in solid rather than drawn as lines.
The rest of the series
- The wave scanner What this loop does once there is paint in front of it.
- The Oramics machine Who Oram was, how the machine came to exist, and what was actually new about it.
- How the machine read a pitch Four film strips holding one decimal digit each, read as binary-coded decimal.
- Drawn sound on film before Oram Avraamov, Sholpo and McLaren, and the point at which Oram departs from them.
Common questions
What is a flying spot scanner?
A cathode ray tube used as a moving light source rather than as a display. A beam of electrons hits the phosphor and makes a single bright dot, and two voltages, one horizontal and one vertical, place that dot anywhere on the screen. Sweep it fast enough and it can scan an image placed in front of the tube, a point at a time, far faster than any mechanical scanner.
What does a photomultiplier do?
It turns a small amount of light into a usable electrical signal. Light knocks an electron off the first plate inside the tube, that electron is accelerated into a second plate and knocks several more off it, and so on down a chain of plates. The RCA 931A used in the Oramics machine has nine of them, so a single photon's worth of input comes out as a measurable current. It is fast enough to follow an audio waveform.
Why use a feedback loop instead of just reading the brightness?
Because what is being read is not a line of known thickness but a thick, irregular ribbon of paint, which has no single value at a given position. A feedback loop sidesteps the problem: the spot is driven upward until light gets through, so it comes to rest on the top edge of whatever was painted. The voltage required to hold it there is the measurement, and the shape of the paint never has to be interpreted.
Sources
Tom Richards, Oramics: Precedents, Technology and Influence, PhD thesis, Goldsmiths, University of London, 2018. Graham Wrench, who built the wave scanner between 1964 and 1966, described the circuits in an interview with Sound on Sound. Written for Drawing Sound.