Interview with Colin Anderson

Hi. I’m here with Colin Anderson, the managing director of Denki, a digital toy workshop in Dundee. Denki lead the way in interactive television gaming. But before he set up Denki, Colin was the head of audio at DMA Designs, the company that gave us Grand Theft Auto and Lemmings. Colin, you were working in game audio at a really exciting time, one that saw a shift away from the sound chips of the 8-bit and 16-bit machines towards sampled sound cards and sound streaming off CD-ROM. How did that change the way that you used sound in video games?

It changed in a lot of different ways, but probably the most significant was – it was pros and cons, it wasn’t just all pros – but the pros that you got was definitely the fidelity of the audio that you could create. It gave you access to the same resources that the film and television industry would use, for example, and that meant for the first time you could use real recordings of real instruments, of real sound effects, so actual doors closing, instead of having to synthesise them in some way approximately. So it was really good from that perspective.

It also gave you great use of talent as well, because suddenly you didn’t have this barrier, where it was only coders who could work in the space. You had this thing where you could work with sound designers who had experience in other media. So that was good. The cons were things like interactivity. We lost interactivity for a while. The synth chips were particularly good because they were being coded at quite a low level. They were really good at responding to gameplay as it moved, and that sort of went away when we started using CD and things like that. And also expectations changed really quickly as well.

Suddenly the novelty of, “hey, this has got a CD soundtrack” went away, and people were just like, OK, we expect that. Of course it’s going to have a CD soundtrack. What else have you got? It was a real game changer in that respect. So, yeah, there were pros and cons. But it was, as you say, a real exciting time. Now, one of my favourite games from that period was the original Grand Theft Auto. Now, in that, you used an in-game car radio station to add an extra layer of game narrative. It’s quite a simple idea, but it’s one that works really well. How did the idea for that come about? It came about during the process of making the game.

So it would be wrong to say that we started with the idea of having radio stations. It was something that evolved in the process. And I think that’s true of a lot of the great ideas that people find, looking back through games that have made a real impact, is that most of the things that people remember usually weren’t there right at the start. They came along as part of the journey, if you like. And the radio stations, in particular, was just a case of we were going through creating snippets of tracks in order to test out what kind of music we would put in this.

Because the plan originally was, first of all, to use MIDI for the soundtrack and use MIDI instruments and, secondly, to have one track that started at the beginning and finished at the end, just like any other game of that particular time would. And it was only when I was reviewing the tracks with Craig Conner, who was writing the sketches at that point – he’d written about 10 or 20 of them – and in the process of going through it, we started to pick up on this. It’s like, these ones, because it was the same person, they had a certain style. So these ones sounded like dance tracks. These ones sounded like techno. These ones sounded like pop, whatever.

And it was really that that inspired the kind of epiphany, where suddenly, when I was listening it, I just went, you know, these would be great if you just separated them out into radio stations. And you saw Craig’s eyes light up, and that was a reflection of me saying it, which I wasn’t really thinking about, and seeing Craig’s reaction to it, that made me pause and think, actually, that would work pretty well, wouldn’t it?

And from there it was just one of those moments where you realise that the idea fitted the context of the game, and that was the thing that really makes the idea special, I think, is not because it is a great idea in and of itself, but it was the perfect idea for that game at that time. But it was only as good an idea as the execution, as the implementation of it. If we had had the idea and implemented it badly, we wouldn’t be talking about it now. It was because we had the idea and then we implemented it really well. That’s what made it memorable.

And presumably, it’s one of those things that you wouldn’t have been able to implement before CD-ROM came along, just because of the volume of audio data that you had to work with to make the radio station function. Kind of. It’s a slightly different way I would look at it. So the actual volume of the track, in terms of space, for example, that is a scalable thing, right? So if we’d been doing it, we could have done the same idea and rendered it with MIDI instruments and with some sort of standard computer synthesis. We could have done that.

It’s not the amount of data that stopped us from doing that, it was purely the fidelity, in that there was no way that you would be able to convince people they were listening to radio stations. It was always going to sound like a facsimile of a radio station. And we were pretty clear that in order to carry that idea off, it needed to be plausible, we needed people to react by going, is that a real band? I haven’t heard this track before. Unless we were getting that kind of reaction from people, we knew it wasn’t good enough. And it was only CD that gave us the ability to kid on that we were real musicians.

And you weren’t just streaming the tracks directly off the CD either, though, were you? You had your own custom sound engine and you would process the music in order for it to sound like it was authentically coming off a car stereo. Could you tell us a wee bit about that? That actually only happened in GTA 2. On GTA 1, it was just coming straight off the CD, and any of the “make it sound like it’s coming off of the radio,” that was all done as part of the recording and mastering process for the tracks.

So by the time they went on the CD, they had a kind of radio sound about them, and we spent quite a bit of time researching the techniques that were used in radio, speaking to people who worked in radio, and finding out what kind of equipment they used. So we did that, but it was really only GTA 2 that we actually started to stream in the raw data and then process that in different ways. But most of that was just down to things like we introduced a range on the transmitters for the stations.

So we actually started to place the stations in the maps, give them a broadcast range, so that it would vary between high quality and low quality as you drove out of range. And the idea was you could kind of triangulate where stations were and locate them in that way. I think – I can’t remember whether we used it in the end, but there was a plan to have one mission where you had to find the radio station that was broadcasting it, nip in, blow it up, and then that radio station would be taken offline. I can’t remember whether we did it now, but that was definitely a plan at one point.

Now, I love that idea that you create a game which has music tracks that people think they might have heard somewhere else, and the radio station really lets you play with that idea. Now, for GTA 2 and GTA 3, you worked with an independent label to produce some of the tracks. They were all original in GTA 1. What sort of balance did you strike between in-house production and commissioned tracks? It changed between the different versions. So in GTA 1, it was 100% completely original material that we recorded in house. There was nothing brought in from the outside at all. And that was a couple of reasons for that.

Firstly, it was because we were this little software development house based in Dundee that nobody had ever heard of really. And at that time, if you approached the record companies and said, would you like to licence us some music for your games, they kind of laughed and said, well, why would we ever want to do that? We’re making so much money from selling you CDs, thank you very much. So it was before that, that whole transition in the music industry. They just weren’t interested. The second thing, though, was more of a creative element, which was it was clearly Grand Theft Auto was an alternative universe.

So it wasn’t a split timeline, where it was meant to share a common timeline with the real world up to a certain point and then diverge, it was clearly a parallel. And because of that, the core of that music radio station thinking was we should only have artists who are not real, because the minute you bring real artists into that space, it’s not a parallel world anymore, it’s suddenly a bit of a confused one. So that was very much the thinking for GTA 1 was we can’t use the real things, so we are going to make it part of the creative design.

I can’t remember which came first, but we certainly took it, embraced it, and ran with it from that point. GTA 2, that started to change because the only people who cared about the music in GTA 1 were the audio team, essentially. Nobody cared about it other than them. Well, that’s not true – the rest of the team – but nobody outside of the team really cared about it. But as soon as it was successful, suddenly people turned up wanting their mate’s band in the game and wanting their favourite tracks to be licenced.

And so it became a bit of a struggle at that point, trying to protect the integrity of that original thought of it’s a parallel world, we shouldn’t have any real artists. So the intention with GTA 2 was we would only allow real artists who recorded new material. That was where we drew the line there. And then by GTA 3, it was clear that this was a losing battle, and I was never going to be able to win that. So we sort of relented. So it increased every time beyond that. Finally with Denki, the style of games that you’re producing now are quite different to the ones you were making 20 years ago.

Is there anything that you learned from your time with DMA Designs that apply equally to the games that you’re making now? Are there are any universals in video game music? I couldn’t do what I do now at Denki without what I learned DMA Design because everything that I did in that company has directly influenced how we’ve gone on to adapt our development methods at Denki. In fact, the reason we set up Denki at the beginning was because I’d identified that most of the things that were holding back interactive audio at that point were nothing to do with deficiencies in the interactive audio creation process, it was all to do with poor methodology and the development practise itself.

So my answer was let’s set up a studio that I can control the development practise and make sure that audio is integrated into the design process right from the start and isn’t just simply expected to be bolted on as an afterthought. And I think you can hear that if you listen to some of the original games that we’ve created over the years, which is Juggle! or Quarrel. They’re clearly experiences that had the audio not been considered an integral part of that product from day one, it just wouldn’t have turned out that way. So we’re not doing anything – we’ve stepped well away from the AAA space, mostly because of the amount of time it takes to produce content in AAA.

I get really impatient. I want to get on to the next thing faster and faster, so a chance to work on projects that take two, three months or a year at most, I’m happy with that. But AAA, you can easily sink three or four years into a single project. And no one’s getting any younger, and we’ve got a lot… I still have lots to learn.

So those elements that I learned in DMA are completely relevant to what we do at Denki, and I would say that if anybody wants to really understand the development process and really wants to understand the limitations of the development process and how it can be made better, then audio is a great place to learn that because your art… you can only do your work and your art if everything else is working, because suddenly, you’re that first discipline who has to pull together all these disparate threads. You have to pull together the code, the design, the visuals, the story, the scripting.

All of these have to come together into creating the perfect audio experience, and if any one of those is broken, you spot it instantly from an audio perspective. I’m not sure there’s any other discipline, possibly with the exception of test, who ever gets to see a project from that perspective, which is why I always say, if you want to know how a project’s doing, don’t ask the producer, ask the audio person.