Live Performance

Every time someone asks me what I do for living and I answer that I make music, the inevitable follow-up question is about what instrument I play. I always get red and mumble "computer", anticipating to spend the next half hour with explanations. (Due to my occupation with Ableton I nowadays simply say that I am a software guy. Usually no further questions.)

How do you play a computer? Rhythmically banging on the screen? With a bow? With a hammer? (sometimes!) The strange look one gets when admitting to play a computer indicates that this instrument does not fit into the typical categories.

A classical non-electronic musical instrument relies on a constant user-interaction in order to produce a sound. The instrument has specific physical properties, defining its sound and the way it wants to be played. The music is a result of theses properties and the skills of the player. The listener has a sense of what goes on, even if they do not play any instrument themselves. It is obvious that a very small instrument sounds different from a very big one, that an orchestra sounds most massive and most complex if everyone there moves a lot, and that hitting a surface with a mallet creates some sort of percussion, with its sound depending on the shape and material.

A whole world of silent movie jokes is based around this universal experience and knowledge. If hitting a head with a pan sounds like a cow bell the comical element is the mismatch between expectation and result. Now explain to someone why pressing a space bar on a computer creates a sounds like a jazz band falling down a stairway one time and the next time it makes no sound at all...

With 'real' instruments it is also obvious that precision, speed, volume, dynamics, richness and variation in sound are the result of hard work, and that becoming a master needs training, education and talent.

The player piano allowed composer Conlon Nancarrow to realise compositions unplayable by humans. The invention of electricity made complex "self playing" instruments possible, and due to further technological progress in electronics and computer science, those machines became small enough to be affordable and sonically rich enough to make them interesting for composers.

If you replace a musician by a sound generating device directly controlled by a score, you get rid of the unpredictable behaviour of a human being and you gain more precise control over the result. A great range of historical computer music and certainly a huge part of the current electronic (dance) music has been produced without the involvement of a musician playing any instrument in realtime. Instead, the composer acts as a controller, a conductor and a system operator, defining which element needs to be placed where on a timeline. This process is of entirely different nature than actually performing music, it is much closer to architecture, painting, sculpting, or engineering.

In electronic music this non-realtime process allows for almost infinite complexity and detail: each part of the composition can be modified again and again. We live in a world of musical undo and versioning. A computer is the perfect tool for these kinds of operations. The workflow in software today is much more efficient than the complex studio setups of the pre-laptop times, which were dominated by a giant mixing desk, lots of hardware units and physical instruments. The result of working for a while on a piece of music with current music software might be a piece of audio which would need hundreds of musicians if performed in one go by human beings. How can this be put on stage?

At the very beginning of computer music, the only way to perform a concert was to play back a tape. Computer-generated music was created in a non-realtime process; the creation of the sound took much longer for the computer than the duration of the sound itself. A situation which, for the design or modification of more complex sounds, still did apply till the end of the 20th century. (In the mid 1990s one had to calculate a simple crossfade in ProTools as a non-real time process!)

As a result, the so-called 'tape concert' was born, and the audience had to accept that a concert means someone is pressing a play button at the beginning and a stop button at the end. Ironically, half a century later, this is what all of us have been experiencing numerous times when someone performs with a laptop. Trying to re-create a complex electronic composition live on stage from scratch is most of the time an impossible and pointless task.

The bottleneck is not that today's computers cannot produce all those layers of sound in realtime, but that one single performer is not able to control the process in a meaningful and expressive way. One cannot perform a symphony alone.

Even back in the 1930s there were already real time electronic instruments, such as the Theremin or Oskar Sala's Mixtur Trautonium. Built for a single player, and by nature in expression and approach similar to acoustic instruments. Their complexity was nevertheless limited and they were never meant to replace a full orchestra. For our purpose of finding ways out of the laptop performance dilemma the tape concert situation is of much more interest, since it is closer to what we do with our laptops today.

Even while these concerts were referred to as tape concerts, there was the notion of the speaker as the instrument. What the audience could see on stage was loudspeakers, and then there was the operator with a mixing desk and the tape machine. The speakers replaced the musicians on stage, whilst the operator was sitting in the middle of the audience or at the back of the room. Visually it was clear that this person was not a musician, but an operator. There was a very practical reason for this. Similar to the role of the conductor, the operator was controlling the sound of the performance, and this could only be done by a placement close enough to the audience. This became even more important once composers started to use multiple speakers.

Multiple speaker tape concerts soon became situations with room for expression by the operator. A whole performance school is based around the concept of the distribution and spatialisation of stereo recordings to multiple, often different sounding speakers, placed all around the listener. The operator, similar to a good DJ, transports the composition from the media to the room by manipulating it. The DJ mixes sources together to stereo, the master of ceremony of a tape concert distributes a stereo signal to multiple speakers dynamically to achieve the most impact. This can be a quite amazing experience, but it certainly needs the operator to be in the eye of the storm, right at the center of the audience.

The DJ concept has a connection to the tape operator concept: Both are taking pre-recorded pieces and re-contextualise them. One by diffusion in space, the other by combining music with other music. When listening to a choir on a record, we do not assume that tiny little singers are having a good time inside a black disk of vinyl.

We judge the DJ by other criteria than the intonation of the micro-choir. A good DJ set offers all we normally expect from a good performance. We can understand the process, we can judge the skills, and we can correlate the musical output to input from the DJ; we have learned to evaluate the skill of working with two turntables and a mixer.

The same is true for the distribution of pre-recorded music to multiple speakers. There is a chance we can understand the concept and this helps to evaluate the quality of the performance more independently from the quality of the piece performed.

Often a classic tape concert comes with an oral introduction or written statement, helping the audience to gain insight into the creation of the presented work. I think this could also work as model for today's presentation of various kinds of electronic music. However, while in the academic music world tape concerts are well accepted and understood, there seems to be a need for electronic music outside that academic context to be "performed live" and "on stage", regardless of whether this is really possible or not.

When listening to a laptop performance, the audience might compare a completely improvised set (which is indeed also possible now with a laptop if we accept reduced complexity of interaction) with a completely pre-recorded set.

The audience cannot really judge the quality of the performance, only the quality of the underlying musical or visual work, they might compare a full playback with a risky improvisation - without being able to distinguish the two. Also the performers might want to be more flexible, might want to interact more.

While the creation of academic computer music was from the very beginning more a non-realtime process than an actual performance, the late 1960s saw the development of the first commercial synthesizers, equipped with a piano keyboard, and ready to be played by musicians rather than operated by engineers. Suddenly, electronic sound became accessible to 'real' musicians - this led not only to impressive stacks of Moog, Oberheim, Korg, Arp, Yamaha and Roland keyboards on stage, but also to a public awareness of electronically created sounds. The peak of this development were the stage shows of electronic post-rock bands like Tangerine Dream and Kraftwerk in the late 1970s and early 1980s.

What the audience saw was musicians operating impressive amounts of technology, blinking lights, large mixing consoles, and shimmering green computer terminals - a showcase of the future - on stage.

It was impossible to understand how all these previously unheard-of new sounds emerged from the machines, but due to the simple fact that there masses of them and total recall was not possible, the stage became a place where a lot of action was needed throughout the course of the performance.

Cables had to be connected and disconnected from modular synthesizers, faders had to be moved, large floppy disks were carefully inserted into equipment most people had never seen before, not to mention owning one of these precious units, and more complex musical lines still had to be played by hand. It was a happening, an event, remarkable and full of technological magic; the concert was a unique chance to experience the creation of electronic music. And in general it really was 'live'. What was set up on stage was nothing less than a complete electronic studio, operated by experts in real time. Great theatre, great pathos, and sometimes even great musical results.

These precious moments come to an end driven by the same forces which were enabling them: Computers became cheaper and more powerful. More functions could be calculated in real time inside a small box.

This development changed electronic live performance in significant ways. The more complex the musical output created with a laptop became, the less possible to re-create it live. A straight techno track made with a Roland TR-808 and some effects and synths can be performed as an endlessly varying work for hours. A mid 1990s drum & bass track, with all its time-stretches, sampling tricks and well-composed breaks is much harder to produce live, and marks pretty much the end of real live performance in most cases. As a result, most live performances became more tape concert-like again, with whole pieces played back triggered by a mouse click and the performer watching the computer doing the work.

This would all be fine, if performance practise were reflecting this, but obviously things did not work out that way. Instead we experienced performers who were more or less pretending to do something essential, or carrying out little manipulations of mostly predefined music. The performers became slaves of their machines, disconnected from their own work as well as from the audience. This also has to do with the second big motor of change: fame.

Fame puts the performer on stage, away from the audience. Miniaturisation puts the orchestra inside the laptop. Fame plus miniaturisation works very effectively as a performance killer.

That intimacy of a small club and the audience very close to the performer provided a highly communicative situation, where interaction with the audience was possible, if not even desired.

But once the artist reaches a certain level of fame, this does not work anymore; The audience gets bigger, and the performer moves on to the big stage. The audience expects from a concert the same full-on listening experience as from records. And this is impossible to deliver in real time. But the artist has not much of a choice. They play more or less pre-recorded music. From a laptop. Far away from the audience. Elevated but lonely.

This situation not only leads the audience to conclude the person on stage might be checking emails or flight dates, but it is also unsatisfying for the performer. Performing electronic music on a stage without acoustic feedback from the room, completely relying on some monitors, is quite a challenge and often enough far from being fun.

Add a sound engineer to that scenario who has no idea at all about your aesthetics, and no band colleagues who could provide some means of social interaction. Instead, there is just you and your laptop. The best recipe to survive this is to play very loud, with very low complexity and hope for an audience in a chemically enhanced mode.

Unfortunately most typical concert situations outside the academic computer music community do not support the idea of playing right in the middle of the audience. In a club, it is often impossible since there is the dance-floor and you do not want to be there with a laptop on a small table at four in the morning. Even if you do find a situation appropriate for a performance in the middle of the room, maybe at a festival, and after successfully arguing with the stage manager for several hours you might be confronted with the dynamics of audience expectations:

There is an interesting difference between the computer music presenter and a live act. While the tape operator in the middle of the room has perfect conditions for presenting a finished work in the most brilliant way (which might occasionally even include virtuoso mixing desk science rather than static adjustment to match room acoustics), the live act has to fight with situations which are far from perfect and at the same time is expected to be more lively. Given these conditions, it is no wonder that generally rough and direct live sets are more enjoyable, while the attempt to reproduce complex studio works on a stage seem more likely to fail.

A rough sounding performance simply seems to match so much more the visual information we get when watching a person behind a laptop. Even if we have no clue about their work, there is a vague idea of how much complexity a single human being can handle.

If we experience more detail and perfection we most likely will suspect we are listening to pre-prepared music. And most of the time we are right with this assumption.

We could come to the conclusion that only simple, rough and direct performances are real performances, forget about complexity and detail and next time we are invited to perform we grab a drum computer, a cheap keyboard, a microphone, and make sure we are really drunk. It might actually work very well.

But what are we going to do if this is not what we want musically?