Warren Cann, Ultravox, and the strange business of teaching machines to make music
In 1980, Ultravox released Vienna, an album which helped define the sound of the coming decade. It combined European romanticism, aggressive synthesiser textures and Midge Ure's voice. It also had Billy Currie, who could make an ARP Odyssey sound as though it were having a nervous breakdown.
Warren Cann, the band's drummer, gets far less credit. He was also a programmer, an electronics enthusiast and an experimental instrument designer. He ran a growing collection of machines which the band christened the Iron Lung.
What happens when a drummer stops only playing an instrument and starts designing the system that makes the music? In Cann's case, the answer runs from early Roland rhythm boxes to programmable drum computers. It ends closer to cybernetics and human-computer interaction than you would expect from a band associated with the New Romantic movement.
The problem with rhythm boxes
By the late 1970s, synthesisers had escaped from experimental studios and progressive rock. They were finding their way into punk, new wave and the emerging electronic pop movement. Electronic percussion lagged behind.
The early rhythm machines were built largely to accompany organists and other performers, providing a selection of preset rhythms. The Roland TR-77, introduced in 1972, was typical. It offered little control over individual notes, no user-programmable sequencer, and no way to recall an exact tempo.
That was adequate for accompanying an organist. It was far less use to a drummer performing complex electronic pop in front of several thousand people.
Cann kept the instrument and had it modified.
Modifying the machine
Cann described his customised percussion equipment to Electronics & Music Maker in April 1981. His TR-77 had separate outputs for the bass drum, snare and hi-hat, so each could be processed on its own. Added controls changed the snare's noise and the bass drum's attack and pitch. His Roland CR-78 gained a bass drum decay control.
The changes had a practical purpose. A kick drum that sounds pleasing at domestic levels can become an uncontrolled boom through a concert sound system. Cann needed to change how it behaved, not only how loud it was.
Tempo was the second problem. Early drum machines had imprecise tempo controls, where a tiny movement produced a large change in speed. A band could not count on playing a song at the same speed two nights running.
The 1981 article records the first fix. Both drum machines had fine-tuning controls added because their original potentiometers were too coarse. An electronic readout, designed by Cann's engineer, displayed the timing setting.
Cann described the system in more detail in a 2010 interview with The Electricity Club. A second potentiometer provided coarse and fine tempo adjustment, while a voltmeter connected to the clock circuit produced a numerical reading corresponding to the machine's speed.
The number did not have to be beats per minute. It only had to be repeatable. If a song required a particular reading, Cann could return to it before the next performance.
In engineering terms, Cann had improved the system's controllability and observability. He could now see a parameter that had been hard to manage, and set it precisely.
The machine must not dictate the music
Cann was not chasing the mechanical precision people associate with electronic music.
In the 1981 interview, he explained that Vienna begins by accelerating, includes a ritardando in its middle section, and returns at another tempo. The machines were not permitted to dictate the pace of the music.
He also described occasionally adjusting the tempo of songs that were nominally intended to maintain a constant speed. His fine control allowed him to do this subtly, without disturbing the rest of the group.
That is a remarkably interesting approach.
Used in the simplest way, a drum machine just obeys. The musician programs a rhythm, presses Start, and the machine plays it back.
Cann's performance system was more complicated. The machine generated a rhythm, the musicians performed in relation to it, and Cann listened to the result and adjusted the machine where necessary.
The machine affected the musicians, and the musicians affected the machine.
Which is where cybernetics comes in.
The cybernetic drummer
Cybernetics studies systems of communication, control and feedback. Its central claim is that you often cannot understand a system by examining its parts alone. You have to study how they interact.
A thermostat is the standard simple example. It measures a room's temperature, compares it with a set value, and adjusts the heating. The heating changes the temperature, which changes what the thermostat does next.
The output becomes part of the next input.
Cann's equipment was not a thermostat, and it would be an exaggeration to say he was consciously applying Norbert Wiener's theories. The analogy still holds.
Cann gave the machine instructions and it produced a rhythm. The band played against that rhythm, and Cann judged the result and altered the machine where needed.
The musical performance was realised and interpreted through an interaction between musicians and equipment.
The feedback was largely human. The drum machine did not listen to the band and adapt its tempo; Cann did.
This distinction matters. Cann had not constructed an autonomous musical intelligence. He had constructed an instrument whose behaviour could be controlled and changed as part of a performance.
The intelligence was in the musician, and in how he worked with the machine.
The technology extended his control over musical time, but he had to invent some of the methods for controlling it.
Musicians would face related questions in the following decades, as sequencers, computers and eventually software instruments became central to music production.
The Trigger Recorder Sequencer
Cann also conceived a custom device called the Trigger Recorder Sequencer, which his engineer built.
Recording studios traditionally used a click track to establish timing. Once recorded, the click track was fixed. If the band later changed or extended the arrangement, synchronising new electronic parts to it became difficult.
The device recorded patterns of trigger pulses in synchronisation with existing material. Individual sections could be erased, changed and recorded again.
Toggle switches directed the resulting pulses to different electronic instruments, including drum machines, a handclap unit and synthesisers.
The system separated the timing of musical events from the sounds those events produced.
That is a significant conceptual distinction.
A trigger pulse does not intrinsically represent a snare drum, a bass drum or a synthesised note. It represents an event occurring at a particular moment. The instrument receiving the pulse determines what happens next.
The same underlying organisation of musical time could therefore control different instruments.
Similar ideas were being explored elsewhere in electronic music. Before MIDI became widely available, however, connecting instruments from different manufacturers was frequently an exercise in improvisation, electronics and considerable patience.
Cann and his engineer were devising their own solutions.
The Trigger Recorder Sequencer was used during the recording of Vienna, including additional work on the title track and Western Promise.
Ultravox's music depended not simply on individual synthesisers and drum machines, but on systems of interconnected instruments which could be controlled, modified and synchronised.
The Iron Lung
The equipment grew, and so did the work of running it.
Over time, the Iron Lung incorporated the Roland TR-77 and CR-78, Linn drum computers, Sequential Circuits Drumtraks, Simmons electronic percussion, effects processors and custom electronics.
It was closer to a small electronic laboratory than a conventional drum kit.
And Cann ran it while drumming.
At any moment he might be playing acoustic drums, triggering electronic sounds, starting programmed patterns, adjusting tempo or managing the synchronisation of other instruments.
The technology had not removed the performer.
It had given him another job.
Anyone who has operated a complicated electronic rig onstage will appreciate the difficulty. Anyone who has tried to debug one while an audience waits will appreciate it considerably more.
This was not merely an issue of making electronic sounds available in a live performance. It was the much harder problem of making those sounds controllable, reliable and expressive.
The Linn revolution
In 1980, Roger Linn introduced the LM-1 Drum Computer, one of the first commercially available programmable drum machines to use digital samples of real percussion instruments.
For many musicians, its most exciting feature was that an electronic instrument could finally sound something like a real drummer.
Cann valued something else.
In an October 1984 interview with Electronics & Music Maker, he explained that the Linn's programmability mattered more to him than the realism of its sounds. He believed its importance would have been much the same even if it had produced entirely analogue drum sounds.
A musician could imagine a rhythmic pattern, enter it, hear the result, revise it and store it.
That changed how rhythm could be composed.
Playing an acoustic kit means executing a rhythm physically, in real time. Programming allows a musician to construct a rhythm incrementally, including patterns that might be difficult or impossible to perform conventionally.
The programmer works on a different timescale from the performer.
A passage can be assembled note by note. Alternatives can be compared. Changes can be made without another complete take.
The machine becomes a medium for musical thought as well as a means of producing sound.
Cann was particularly enthusiastic about the Sequential Circuits Drumtraks, which allowed tuning and levels to be programmed alongside the rhythmic sequence.
Each development increased the number of musical decisions that could be represented, stored and changed.
For Cann, that counted for more than realism.
Composition as programming
It is tempting to treat drum programming as a lesser activity than drumming, since pressing buttons looks nothing like the physical skill of playing a complicated rhythm.
That confuses two kinds of expertise.
A drummer develops the motor skills to realise musical ideas through physical action.
A programmer develops the ability to conceive, represent and manipulate rhythmic structures through an abstract interface.
Neither guarantees good music.
A technically exceptional drummer can play something tedious, and a programmer with little conventional instrumental technique can construct something rhythmically extraordinary.
Cann recognised this early.
Programmable instruments allowed a kind of composition that did not depend entirely on the performer's immediate physical abilities.
That did not diminish the value of his own playing. Cann remained a formidable drummer.
He simply refused to choose between the two.
He wanted the physical expression of the drummer and the structural possibilities of the machine.
And there is a further consequence worth considering.
The interface of an instrument influences the music that can be made with it.
A drum kit encourages patterns that emerge naturally from the movement of two hands and two feet. A step sequencer encourages patterns that emerge from the arrangement of events on a grid.
Neither is inherently superior. But they offer different ways of thinking about rhythm.
The musician's ideas influence the machine's programming, while the possibilities and limitations of the machine influence the ideas the musician develops.
That is a creative feedback loop, extending beyond the immediate mechanics of live performance.
Listening to the evidence
The best way to understand Cann's contribution is to listen carefully to the records. These five pieces offer different examples of the relationship between acoustic percussion, programmed rhythm and electronic instruments.
Astradyne — Vienna (1980)
The opening track of Vienna is an extraordinary instrumental introduction to the new Ultravox.
Listen to the metallic percussion sounds and the way they sit alongside the band's conventional playing.
In the 1981 Electronics & Music Maker interview, Cann identifies the metallic sound as the CR-78's Metallic Beat.
What's striking is that the sound does not need to resemble an acoustic cymbal. It works because it has its own place in the arrangement.
The drum machine is contributing a distinctive instrumental voice, not merely imitating a drummer.
Mr. X — Vienna (1980)
Here the distinction between electronic and acoustic percussion becomes an interesting listening exercise.
Cann's 1981 interviewer specifically recommended Mr. X as an example of how effectively the two could be combined.
Listen to the individual percussive sounds, then to the rhythmic arrangement as a whole.
The significant achievement is not that electronic percussion is indistinguishable from acoustic percussion. It is that the distinction becomes musically unimportant.
Vienna — Vienna (1980)
Listen to the tempo, rather than merely the sounds.
According to Cann's own account, the opening section accelerates, the middle section slows, and the song returns at a different pace.
Those changes are part of the composition's dramatic structure.
The crucial point is that the electronic percussion participates in those changes rather than forcing the musicians to maintain a mechanically constant tempo.
This is perhaps the clearest recorded illustration of Cann's insistence that the machines must serve the music.
The Thin Wall — Rage in Eden (1981)
By the time of Rage in Eden, the relationship between rhythmic programming and synthesiser sequencing had become even more important.
Listen particularly to the interaction between the percussion and the synthesised bass figures.
In 1984, the band identified The Thin Wall and All Stood Still as examples of arrangements shaped by the combination of acoustic drums and triggered synthesiser bass.
Here the rhythmic structure extends beyond the drum kit.
Cann's role in organising musical time reaches into the parts being played by other instruments.
All Stood Still — Vienna (1980)
This is perhaps the most satisfying demonstration of the principle.
The rhythmic drive depends on the interaction between acoustic percussion, electronic elements and synthesised bass.
Rather than hearing the drum machine as an accompaniment, try listening to the various parts as components of a single rhythmic system.
Each contributes something different, but the arrangement depends on their integration.
The distinction between a drummer, a drum programmer and someone controlling electronic instruments becomes rather less clear.
Which, of course, is precisely the point.
A musician, not a machine operator
Histories of electronic pop tend to concentrate on synthesisers: the instruments, their sounds, and players such as Billy Currie who coaxed remarkable performances from them.
Electronic percussion mattered just as much.
Cann did more than use drum machines earlier or more imaginatively than many of his contemporaries.
He treated them as instruments whose limitations could be challenged, whose interfaces could be improved, and whose relationship with human performance could keep changing.
He did not accept the machine as the manufacturer delivered it.
He wanted to discover what it could do once those limitations were removed.
He also understood that the range of musical ideas a programmable instrument makes possible can matter more than the sound it produces.
That is why the cybernetics analogy is useful.
Electronic music history is partly a history of musicians adopting new machines.
It is also a history of musicians adapting those machines, discovering new possibilities through interaction with them, and allowing those discoveries to change the music.
The musician changes the machine.
The machine changes the musician.
Warren Cann was working this way more than forty years ago, from behind the drum kit.
And perhaps it's time we paid a little more attention to the drummer.
Sources and further reading
-
Mike Beecher, “Ultravox”, Electronics & Music Maker, April 1981. Contemporary documentation of Cann's modified instruments, tempo controls, percussion equipment and Trigger Recorder Sequencer.
-
Dan Goldstein, “Love's Great Adventure”, Electronics & Music Maker, October 1984. Cann's reflections on drum programming and the evolution of electronic percussion.
-
Chi Ming Lai, “Warren Cann Interview”, The Electricity Club, 13 October 2010. Retrospective technical details on the Iron Lung and modified Roland equipment.
-
Warren Cann, “The Linn LM-1 Drum Computer”, Electronics & Music Maker, September 1981. Cann's contemporary review of the instrument.
-
Jonas Wårstad, “Ultravox: The Story”, interview material at discog.info. Cann's recollections of the band's early equipment and collaborators.
-
Norbert Wiener, Cybernetics: Or Control and Communication in the Animal and the Machine (1948). Background to the theory of feedback and control systems.
- You can also read my previous blog, “In Praise of Ultravox“, in which I looked at their four studio albums released between 1980 and 1984, picking out the highlights and explaining why I enjoy them so much.