Just finished up a deeply dispiriting session in which I tried to produce some music effects for GuruQuest in the continuing tradition of doing everything for the game myself. The basic idea was sound, I think, but the execution and everything else was oh so disappointing.
The plan was to record chords from the C major progression and then have the game play them at some kind of random interval during play to create a bit of a mood and, most importantly really, just a little bit of sound texture to an otherwise soundless world. Anyway, in my mind at least, this would create a series of sounds that would be compatible with each other, but also randomly selected, in keeping with the “generated” idea behind the game as a whole.
But instead I got a big taste of the experience of wading into into the unknown realms that stretch as far as the eye can see outside my limited expertise. See, I can do some programming, and on a good day I’ve got an alright eye for visual aesthetics, but on the sound front? Well, I can plink away at a ukulele and I know the bare minimum about sound programming, and that’s that.
Recording music is really hard. I guess that’s why there are highly paid professionals and, importantly, special rooms. My setup was a noise cancelling mic (tuned for human speech presumably), my laptop (with its fan noisily running because it’s an old fellah these days), and my ukulele. This did not turn out to be the best audio setup one could hope for. But I forged on and learned various things about the proximity of the instrument to the microphone and so on. Then I learned how the noise-cancellation algorithm in Audacity (the sound editing software) can make the ukulele sound bizarrely terrible. So that wasn’t great.
After many painful run-throughs, I ended up with sound files for each of the chords in the progression. They had a bit of background hiss, but not so much to be totally noticable, so I wasn’t completely unhappy. Then I tried installing some sounds in the game world so that a random chord would play every four seconds or so.
And then I was completely unhappy.
It sounded so very terrible. Not so much because of the sound quality, which was passable, but because the pairing of the (medium speed) chords and the still, clean aesthetic of the game was wretched. I don’t think I’ve felt such a jarring dislike of something I’ve made in a long time.
Of course, the flip side of “oh god, that’s awful!” is a heapin’ helpin’ of learning. Clearly if the ukulele thing is going to work the chords need to be much slower (more like individual notes) and much softer. It might stand a chance of working. So I’ll have to re-record it all and keep my chin up and so on. And, at the very least, beneath the sadness and despair, at least I did the thing I was thinking of and found out it sucked, rather than blithely assuming it would be fine later, later.
Anyway, much respect to all you proper music and audio people out there – I very literally don’t know how you do it.
I’ve been trying to get back into playing around with Processing again, after my quite productive jag making webcam-based platformers a little while back. This time I wanted to tackle some of the sound capabilities thanks to a few student in my class who’ve been talking about their ideas.
Turns out that, for me at least, dealing with sound is a lot harder than video. As I try to understand the library you use (called Minim), I come to realise just how little I know about sound processing jargon and metaphors and technicals. As such, I basically don’t understand how anything works except by tweaking example code. Further, my lack of understanding pretty much prevents me from tweaking the code in interesting and/or useful ways. I’m a bit stymied.
To this point the best I’ve managed is to get some live waveforms to appear, with the idea being that I could get them to behave as landscapes that you could then walk on in a game. I have a bunch of (I think) cool ideas that could go with that basic starting principle, but my lack of ability to comprehend how the sound stuff in Processing works is really holding me back. I just don’t think in audio.
Anyway, I’ll do some more battle with it if only because it’s weirdly fun testing the crappy prototypes by singing and clapping and beatboxing into my macbook’s microphone. In private, of course.