[Note: I'm having a little problem getting the music samples to play first time. If you experience the same, try hitting pause and then play again - that fixes it for me]

I often see an interest in the composer's process, how they get their ideas and how they then turn it into a final version and how it morphs and changes over the process. The focus in this piece will definitely be more on the mysterious, high level processes that occur around coming up with musical ideas and iterating on them, rather than technical details.

The complicated answer is that it very much varies. Ideas for music (like anything) can start anywhere. It could start as a rhythmic idea or a harmonic idea, or a desire to play with a certain type of sound, and from there it can spin off in various directions. Still, I thought it might be useful to show you one particular example from start to (almost) finish, to show you how it might work in that case.

I was approached about doing the soundtrack for Mimic, a local-multiplayer party game about trying to blend in with a crowd while identifying the other people trying to hide in it. 

The current title screen of  Mimic

The current title screen of Mimic

Our initial conversations were around the tone of the game, and how the music should feel. since there's something just a little B-movie about the aesthetic of the game, points of reference quickly became Mars Attacks! and especially that level in Psychonauts with the conspiracy theory neighbourhood. 

My early thoughts about the soundtrack for the game were to make heavy use of theremin, so the little music inventor that tumbles around in my head and tosses musical ideas out had its keyboard very much set to 'theremin' in the beginning. That little music inventor is a tricksome fellow, and won't necessarily throw out an idea on demand - more likely  it'll wait until I'm half-distracted by something else and sneak an idea in there. This doesn't feel like part of the process that I'm very much in control of, but it's vital.

I have my best ideas while walking or in the shower or doing some other activity where I'm not simply sitting down trying to generate ideas, but I'm moving about and slightly distracted, so journeys to and from work and even when at home simply getting up and pacing can be important.

In this case, I hit upon a motif that would eventually become the seed for Mimic while in a train station. Not wanting to lose it, I hummed it quietly into my phone, just so there would be enough to jog my memory later when I could get to my laptop.

The next step was to turn this into a couple of musical ideas that would be expressions of this motif, purely as exploration, and to give an idea of where I was going thematically with it. This resulted in two 'sketches'.

The first sketch was the most sketchy - I wanted to see if there was some value in using a similar approach Peter McConnell used in the Psychonauts clip, with various overlapping instruments creating a creepy, paranoid atmosphere but fairly loose rhythmically and tonally. 

This was fun, but it didn't feel like a direction for a time-based party game, even one with a relatively sullen atmosphere in comparison to most. I wondered how it might go if I went totally in the other direction and treated the tune as a big film theme, Danny Elfman style complete with strident horn accompaniment.

On a briefly-more-technical note, this was a learning process, since although I use orchestral sounds regularly in my work, I've never really gone for something that sounded quite so big. In making this sketch I learned quite a lot about creating a bigger sound from a limited sample set. Compare the first version here, to the second - the instruments used are identical, but the way they are mixed and compressed to make more space for them makes a world of difference:

Making this sketch was a blast, and I have a feeling that it might make its way into the game somewhere or other, but it still didn't feel like what the game needed. The title screen was more subdued, and what I was  starting to think about by this point was how the music might work during play.

Mimic In-Game, Work-In-Progress screenshot.

Mimic In-Game, Work-In-Progress screenshot.

A conversation about that made me start thinking about music with multiple layers - a basic bed that would have more parts added to it as the level came closer towards a close. With that in mind, I set about taking the musical ideas I'd explored so far and making something that might actually work in play.

This next sample is my first draft of that work. At this point I was still attempting to focus on an orchestral sound, with a prominent theremin lead. The sample was inteneded to show how a (at this point obvious) loop might start to bring in more instruments to become busier as the piece goes on. 

These examples get quite lengthy from now on, so you may want to skip around a little using the interface to get the general idea. 

I was very please with how this came out, on the whole. I think it successfully added a much needed feeling of playful urgency. When I was writing I had a feeling that the mood I was conveying was inspired by something but it was only later, watching Indiana Jones and the Last Crusade that I realised I was probably trying to evoke a similar sort of playful, comedic tension as the music underscores in this scene:

It seems appropriate that my music for hiding in disguise amongst a crowd of people was unconsciously inspired by a film scene about hiding in disguise amid amongst a crowd of people!

Still, there was something missing. What I was feeling was a disconnect between the onscreen visuals and what I was hearing. The music was a touch too lush, a bit too natural, where the art has a synthetic, pixelly look which wasn't really coming across in the music.

After taking some time to give this a lot of thought, I returned with new ideas about the instrumentation, replacing the military style snare drums with a morphing, squelchy electronic beat and augmenting the pizzicato strings with electronic bass as the piece intensifies. I also doubled up the theremin part, de-emphasising it and  coupling it with a more modern sounding, bell-like synth. 

This produced a pleasing mix of natural and unnatural sounds that finally cemented the sound-world of Mimic in my head, and here is a closer-to-complete version of the standard level music as it builds.

Relieved at finding the sound I wanted, I revisited the title screen, looking for something with the same motif that was more subdued as befits the dusky cityscape depicted, with the crowd below. From this early sketch, it looks like the theremin might get a chance to shine after all:

Armed with a better understanding of how I want the world of Mimic to sound, iteration should be simpler from now on. 

AuthorPeter Silk