Writing my review of the multimedia elements of the Tate E-learning portal reminded me of another review I wrote seven years ago. Back then CD-ROMs were still seen as new and a bit experimental. I thought there was a rich vein to be mined at the point where new multimedia 'toy' and game interfaces were used to make/manipulate/re-mix music. I thought the Header CD-ROM from Tui Interactive Media was the best example I'd seen of this at the time.
I hoped to stir up some of my comrades in the Usability/Human-Computer Interaction professions by championing something quite different from the interactive 'systems' we were used to working on. I failed in that and my other objective, but before I go into the details, and how things have changed in this field since 1996, here's the historical article, originally published in the British HCI Group's Interfaces magazine.
Review of Header #1 CD-ROM, September 1996
This CD-ROM challenges much of the received wisdom about usability and multimedia. From the opening screen — where the menu choices swirl in orbit and depth of focus, challenging you and the cursor to 'catch' one of them — it confounds conventions of user control and user feedback. This can be a risky tactic, and many recent multimedia titles have sacrificed even basic usability in the name of innovation. But with Header the risks mostly pay off. In the process, it presents its musical content in a range of settings which subtly reframe our ideas of what recorded music is, and what interaction with music can be.
Header is an independently produced 'enhanced CD' combining 30 minutes of standard audio CD with CD-ROM multimedia. The ROM section is twelve tracks presenting different techniques for de-constructing and re-constructing musical pieces from dance and reggae artists, plus occasional interviews in Quicktime video clips. Apart from the location of the two basic navigation buttons, each track has a distinct look and feel, so users have to feel their way, exploring different mouse movements and trying to monitor their effects. As some of the effects are quite subtle, and some changes occur spontaneously, this is not always straightforward.
On one piece the relative position of different elements in the mix (drums, bass, etc) is dependent on the distance of the cursor from the icons for each element; you can also 'pick up' the icons to bring different combinations closer together or further apart. On another piece, five different tracks of music are arranged side by side: you hear whichever one the cursor is over, but there is a delay built into the system so that you can segue different tracks in different ways according to how fast you move the cursor. In a third case, different musical elements jump spontaneously around a square grid, but you can use the cursor to turn them 'on' or 'off', or to divert their movement.
This is not the first multimedia application in the creative media field to break a few rules in the human-computer interaction guide book. But it's the first I've seen to do so knowingly, and to go back to the basic grammar of interaction, before the guide book became (prematurely) a tablet of stone.
Header reminds me in places of Myron Krueger's early demonstrations of 'artificial reality'. These explored worlds where the computer-generated environment might interact with you (or your avatar) without you asking it to, and where objects acted differently — sometimes unpredictably — according to the distance or overlap between them. This was in the 70s, before task analysis and other social science techniques framed our perception of what people do with computers. It was before pointing and clicking defined computers as essentially passive tools and information spaces for the user to navigate around like a utilitarian magpie.
Header takes us back to Krueger's more elementary style — albeit now with rich graphics — and applies some of these early interaction techniques to the substance of music itself. It breaks down Carl Craig's Sonic Distortion track into five component elements and represents them as blue circles in constant motion. Passing the cursor over a circle turns that element - the bass track, or the vocal, say — on or off, thus allowing you to 'mix' the track in different ways. Or you can leave the cursor still and let the movement of the circles determine its own mix. And the audio part of the CD lets you hear Craig's own mix — implicitly no longer 'definitive' — for comparison.
It seems fairly clear now that the future of interactive media, for mass audiences, is not in turning everyone into a director-cum-editor or a producer. Choosing camera angles and experimenting with different treatment effects is hard work. Not something you can do when listening in the car, on your Walkman, or while doing the ironing.
Both the 'choosing alternate versions' and the 'hypertext encyclopaedia' paradigms of interactivity remain aloof from their media sources and keep these sources intact except at the broadest level. Like the Truth in the X Files, the media content is 'out there': where you can surf but not really engage. In contrast, some of the pieces on Header bring you right up close to the music. On the Mo Wax piece you can grab the sound track with the mouse and move it like a scratch record deck. On the Unkle piece different tracks can be found, iconised, adrift on a background that slides across the screen at a speed and direction determined by the location of the cursor. Their level in the mix is determined by the distance of their icon from the cursor. So as the graphic landscape glides past, the tracks fade in and out seamlessly.
This representation of the source music content is sometimes more physical and sometimes more abstract than we are used to. As opposed to the abstraction that turns music into staves in traditional notation, the Header abstractions — into space and graphic effects — have the potential to create new synaesthetic effects for the user.
On a purely musical level Header offers ways of experiencing that you probably haven't had before. Four Horace Andy tracks can be manipulated, by moving the cursor left-right, to make them more or less 'dubby'. The argument against interactive music is that, if you like a track in the first place, why would you want to mess with it? (And if you don't like it, what's the motivation to mess with it?) This piece gives you a way of messing with the music — but not too much — in a way that is totally in sympathy with the 'original' source. The discovery of this effect delighted me, and hasn't worn off quickly. I could happily have a thumbwheel attached to my steering wheel to add more dub to some of my tape collection while driving down the M1.
The range of interaction techniques and effects on Header make exploration and discovery a key part of the experience. It is more integral than just 'Can I click here, and what happens if I do?'. And this is where Header challenges the orthodoxy of usability: the musical and graphical environments are often active from the start, as in the Carl Craig track for example, and it is not immediately apparent if or how you can affect them. It takes a while to orient yourself and experiment and assess the impact of your experiments, which may be complex or subtle.
It doesn't always work: I still don't know what difference it makes whether any or all of Carl Craig's blue circles in the left or right section of the screen. Nevertheless, the necessity of an exploratory attitude to the interface is critical to its overall effect. There aren't many CD-ROMs which still have new facets to reveal — like a musical instrument does — after months of playing.
Part of the success of Header depends on its choice of musical sources. Dub reggae and techno dance music lend themselves to exploratory musical manipulations more readily than, say, Oasis. Similarly, the DIY mixes in the ROM part of David Thomas's latest CD, Erewhon, (which bears some similarities to Header) get away with sounding weird and not-quite musical, because David Thomas's songs have always sounded weird and not-quite musical! But there is still a rich field that has been opened up here.
Of course Header has its disappointments. As with other multimedia 'magazines' like Blender, the short video clip seems the least satisfactory of all means of presenting interviews. There are also a number of 'filler' pieces, such as the As One piece which, in its static presentation of several tracks for mixing together by waving the cursor over them, is functionally the same as less inspired multimedia mixing simulations like Mixman.
Header works as a magazine, rather than an album. That is: it is something I find myself more likely to pick up and flick through than to sit down and really develop a relationship with (though when was the last time I did that with an album?). Nevertheless it is rich and engaging enough to sustain repeated explorations of the same material — which is not something you can say of many CD-ROMs. As a sampler of an index of possibilities, it is certainly enticing. I would love to see what these designers could do with a larger canvas than half a CD-ROM affords. In the meantime, I'm placing my order for Header #2.
Krueger, M (1990) Artificial Reality. Addison-Wesley.
That penultimate sentence about the 'larger campus' refers to a second thinly-veiled objective I had: to persuade my friend Tim, then creative director of the National Centre for Popular Music, to commission the makers of Header to make some more ambitious multimedia pieces for the NCPM (sadly now defunct). I failed there as well.
What has happened to the kind of multimedia interfaces to music making that I was so keen on?
The nucleus of the idea at the heart of many of the Header interfaces has become a mini-genre of its own, known as 'soundtoys' and with its main collection at http://www.soundtoys.net. The difference is that these toys rarely use 'established' or well-known music tracks. I guess people concluded that the business equations just didn't work: adding a new interface to the music as part of an 'enhanced CD' probably cost more but didn't generate many extra sales; conversely, the cost of obtaining rights to well-known music could not be recovered by sales of multimedia titles.
The 'toy' terminology shows the modest claims of this approach now. In 1996 I, and others, hoped these new interfaces to music making might realise a grander ambition.Posted by David Jennings in section(s) Human-Computer Interaction, Music and Multimedia, Reviews on 5 September 02003 | TrackBack