Skip to 0 minutes and 9 seconds It took developers time to explore the intricacies of the SID and its idiosyncrasies, but even in the very earliest games, you can see the influence of the popular music of the day shaping the game soundtracks. Musically, David Whittaker, the coder behind the early Commodore hit, Lazy Jones, was a New Romantic, long before he got into video game music. The New Romantics, in contrast to the aggressive austerity of punk, built their music around a synthetic sound that was initially, at any rate, quite eerie and sterile, which was at least in part a consequence of the musicians exploring the capabilities of the digital synths that were entering the market in the early 1980s.
Skip to 0 minutes and 55 seconds Over time, dance-infused rhythms and arpeggiated synth sequences were incorporated into the music, and by the mid-1980s, the British synth pop that had grown out of New Romanticism had a warmer, punchier sound. And that was the sound that Whittaker adopted for his best known soundtrack, Lazy Jones.
Skip to 1 minute and 31 seconds The music uses only two of the three voices of the SID chip. The third was reserved for sound effects. The music is simple but effective – strong melodies supported by a bouncing bassline, played in broken octaves.
Skip to 1 minute and 50 seconds That bassline provides a sense of continuity, and a dynamic movement to what might otherwise be quite a static, synthetic track. Already a staple of 1980s-era synthpop – think New Order’s Blue Monday – it became one of the key characteristics, not just of C64 music, but of 8-bit video game sound in general. As developers began to unlock the capabilities of the SID, the musical data and its timing became more and more complex. Most SID musicians wrote custom code, a sound driver in assembly language to manage it all.
Skip to 2 minutes and 32 seconds Now ostensibly, sound drivers all do exactly the same thing: they take musical data, and they use it to manipulate the SID’s registers. But just as the touch of different musicians can sculpt very different performances from the same instrument, the code of different computer musicians could impart very different sounds from the same hardware. Take vibrato, for example, that slight pitch modulation that gives a sound movement. There are all sorts of unique characteristics to it. There’s the delay between the note and the vibrato onset. There’s the vibrato depth, the rate, and so on. All of that detail, that performance technique, is coded in the driver. And everybody did it definitely.
Skip to 3 minutes and 24 seconds But of all the game composers of that time, perhaps the best known, and the one who most skillfully combined an innate sense of musicality and coding prowess was Rob Hubbard. At a time when the industry was still professionalising, he brought the work ethic of a professional musician. I met up with Rob, and asked him how writing music in code rather than on manuscript paper might have changed his approach to music-making. Rob. Hi. How did you get started in video game composition? I was working as a musician, doing plenty of gigs, and at the time, you know, all the press were saying, if you’re a musician, you need to start learning about computers, because that’s going to be very important.
Skip to 4 minutes and 14 seconds And so that’s what I did. And I just learned how to program. And I looked at educational software, and that was basically a nonstarter, so then I looked at the games side of it, and that’s how I got into doing the music for games. It wasn’t quite as straightforward as that, but essentially, that’s what happened. So you were already quite an accomplished professional musician before you started writing video game music. Those professional musicianship skills – did they come in useful? Yeah, I think that when you start learning an instrument as a kid, then you develop your skills as a teenager, and then work as a professional musician, what you learn is discipline.
Skip to 5 minutes and 1 second You have to be very focused and disciplined to learn an instrument in the first place. And if you work as a musician, you have to be very professional about what you do, otherwise you might not keep the job, you know? So when I got asked to do stuff, they wanted it at a certain time and everything else, it was always right there, it was always delivered, no matter how much work you had to do. So I think that professionalism and discipline really came in very useful. And could you describe what your process of working was like? How easy was it to convert your musical ideas into machine code?
Skip to 5 minutes and 47 seconds It was actually quite tedious, quite difficult – the process involved for me… I used to write sketches on paper, and then I would code them up in assembly language – machine code – which is all a lot of hexadecimal numbers, and then I would basically work on maybe two to four bars at a time, and I was able to edit synthesiser parameters and I could edit the music in real time and try to get it to sound the best way I could before moving on, and then develop the composition that way.
Skip to 6 minutes and 30 seconds It was never a case of writing out six minutes worth of music on manuscript paper, and then coding the whole thing up and then trying to hope that you hadn’t made any mistakes. So even though you were working on a computer, there was still an element of kind of performance about it? A bit like a synth player twiddling with knobs? Yeah, exactly. Instead of twiddling with the knobs like you would on a Moog or something like that, I was twiddling with numbers in real time. You know, changing things to the absolute maximum I could squeeze out with the little SID chip that was in the C64. Changing things from 1C to 1B, then to 1E, until…
Skip to 7 minutes and 16 seconds Then I would adjust things like the cutoff point and the note length for the release to try to sound right. All these kind of things. You fine-tune it, just to the n-th degree, as much as you could. And you mentioned the SID chip. So through a combination of the SID hardware, your coding, and your musicianship, you were able to write some very sophisticated music. And I can hear elements of maybe Larry Fast or Philip Glass coming through in your music. What was it about the C64 that allowed you to explore that sort of experimental compositional landscape?
Skip to 7 minutes and 55 seconds I think it was a combination of all those things, but it was mainly the software aspect of it, the fact that you had complete control of what was going on in terms of the structure of the music with the software is very easy to get into those kind of more minimalistic type of styles, which were also popular in the mid-80s. There was quite a strong – you know, people like Kraftwerk, people like Jean Michel Jarre. Larry Fast is somebody I stumbled across, well, stumbled upon. And there was a kind of – on the pop side of it, there was a New Romantic kind of movement with all the synth bands. So that kind of played into it as well.
Skip to 8 minutes and 47 seconds So the synthesiser element was very strong. On the compositional side, the software allowed you to kind of explore some of those other possibilities. I mean, part of my dreams that I had at that time was, if I had, like, two Prophet V synthesisers that I could control with a microprocessor, then that would have been the ultimate, you know. So as well as innovating musically, you were able to innovate technically too. Back in the 1980s, you were exploring ideas around adaptive music and interactive music with things like the MIX-E-LOAD system. Could you tell us anything about that? One of the things that happened was through the software, you could control all sorts of synth parameters like pulse width modulation.
Skip to 9 minutes and 38 seconds You can control just about any aspect on the SID chip. But also, when you’re writing code to control the actual structure of what the music is, it then became quite apparent that you could then take that to the next level, and then try to come up with ideas to try to make it more interactive with what was going on with graphics. At the time I was starting to think about the idea of probability-based type of things you could do, even within on a C64. Where you could get a complex kind of bassline or something, and then add like assignments to the notes of probabilities that they would trigger.
Skip to 10 minutes and 31 seconds And then based upon an input from somewhere else, you would then get various degrees of intensity of what that was going to produce. And then not only that, you could then control certain other aspects like harmony, and what kind of sound you would get with a lead instrument, the melody, where, you know, things like that you could then do more interactively. And by the end of its life, you were really pushing the limits of what the Commodore 64 could do musically. At one point, you even made it sound like Jimi Hendrix. How were you able to squeeze those extra voices? People were being able to start doing some real primitive speech things on the C64.
Skip to 11 minutes and 19 seconds So I started to explore how all that stuff kind of worked. And then once I figured out what was going on, they were using a fast non-maskable interrupt running at, I think, about the maximum you get out of it was 5 or 6 kilohertz, which is about 5,000, 6 times – five or six thousand times a second. By manipulating the volume register, which was only four bits, you could get four bit samples running at about 5 or 6K, which seemed ludicrous at the time.
Skip to 12 minutes and 0 seconds But it was a way of trying to take that sound to the next level, and it worked really well with a heavy kind of distorted guitar sample, which I managed to get from a friend of mine who had an old Akai sampler. And also, you know, I could digitise some organ chords, and then try to get those to loop at four bit samples at five kilohertz. And then by changing the speed of that interrupt, you could then chain – you know, get that old – we used to call it munchkin-izing, where the samples are just… You munchkin-ize down when you take the sample rate down.
Skip to 12 minutes and 42 seconds And you know, sometimes the eeriest noise I could get from the samples kind of added to the sound a little bit. But it was a real juggling act to try to get it all to work, because the volume register then kind of screwed up some of the synth sounds that were going on. So you would have to just work like crazy to try to get it all to work and sound reasonable. But the idea of changing the pitch of a guitar sample or other waveforms that I used to do was a thing that was unique at the time. Rob, it’s been a real pleasure. Thank you. OK. Thanks a lot.
Commodore 64: The Music
The SID chip was complex, and Commodore’s implementation of BASIC made it very difficult for new users to work with it.
Even seasoned developers needed time to explore all of its intricacies and it took time to unlock its full potential. In this video, we explore the practice of music-making on the C64, and hear from video game legend Rob Hubbard, who did more than anyone to create the C64 sound.
© 2017, Kenny McAlpine