Hey! Thanks for reading! Just a reminder that I wrote this some years ago, and may have much more complicated feelings about this topic than I did when I wrote it. Happy to elaborate, feel free to reach out to me! 😄
🎵 The song for this post is The Spinach Rag, by Nobuo Uematsu for Final Fantasy VI. 🎵
I was having fun with friends a few months ago where I got to stretch my old sound designer muscles and made three little pieces. I'll first document what they are, then talk a little about how I built them, for if you ever want to try something similar. All were made under pretty severe time and energy constraints, so they're not at "professional" level polish, but I still like how they came out 😊
Personal Climate Impact Sound Effects
The most bite-sized and immediately entertaining, I designed six effects for an installation that told a user if their personal climate impact was responsible or irresponsible. They stood in front of the installation, pressed a button, and were evalutated, playing one of these 6 at random.
I designed a soundscape that played in a set piece of an installation. It imagined the world ravaged by climate change: virtually all animals were dead, and all the vegetation was represented as being contained in the greenhouse. It was supposed to be a peaceful space for people to relax and ruminate, and I produced a generative track that dynamically created a soundscape within parameters, continuously. Here's about two minutes of it.
There's a fair bit of randomness involved in exactly how it sounds; that iteration is a little darker; further down the track, it might sound like this:
Fake Animal Facts
In the same setting above, we intended to have a table with headphones someone could put on to hear looped "facts" about animals that were reconstructed by people desperate to know about them, but completely wrong about everything. It was Attenborough-esque, and meant to loop continously, so anyone could come and put them on to listen to as many or few as they liked.
Most of these were recorded in a panic, by speaking directly into my laptop, in a hallway with people coming and going. Especially in the early "facts," you can hear things happening in the background, and some amateur mistakes like popping in the mic.
I condensed the lines a little so you can hear it all in a few minutes, instead of the 5 meant for passers-by.
All of these were on my Mac laptop, so I keep in mind it's pretty OS specific.
Personal Climate Impact Sound Effects
This required 3 tools. The easiest was the
say command, which lets you use
Apple's historically excellent accessibility skills to make your computer speak
out loud. For example, if you want to recreate the old "DROID" sound from
Android phone commercials, you can run
say -v Cellos DROID, like so.
One of the options lets you save the playback as a file, which is how I made the Samantha voice say all those things.
The next step was sourcing the other sound effects (All Star, Yakety Sax, the dial-up modem noise). This pipeline of tools is super useful (and there are probably other ways to do this).
The first step is downloading youtubedown.pl, a 6000-line Perl script by Jamie Zawinski. It's fantastic, old-internet hackery that takes a YouTube or Vimeo URL and pulls the video down to your computer. Find the sound effects you need on YouTube and download them locally.
From here, I used ffmpeg to extract the sound into its own track.
Lastly, I used the hideous-but-powerful Audacity to split and splice the tracks together, then exported them to a WAV file.
For this I wanted to use generative sounds since it'd be looping in a space for hours and hours. I used Sonic Pi for this, which is a wonderful music coding environment I recommend to everyone to play with; the tutorial has you making sounds pretty quickly and gives you music-making tools in hours, and it's super easy to share and edit the clips of others.
So there were two main components: the chords, and the samples.
Samples are concrete sound effects taken from other pre-recorded media, in this case, animal sounds. I ripped frog, bird, and whale sounds from YouTube, used Audacity to isolate them into individual sounds for the files. Then made a little Sonic Pi script to load them into memory, then loop forever, playing them with little delays.
The chords came from one of the provided examples "Cloud Beat." It sounds like this:
I kept and isolate the code bits that did those 8 second chords (so removed the vinyl hiss and snare):
Then I altered the chords: I asked Karen "what's your favorite musical key?" She said C Minor, so I made that my 1. I added chords for IV, V, VII. Finally, made all the following change at random:
- Which chord in the progression we play (I, IV, V, VII).
- The synth we use (Cloud beat has one, I swap around 5 or 8).
- How "wide" our chords are (five, seven, or nine notes).
- The "character" of the chords (sometimes it plays them diminished, or augmented).
- Ocassionally, we modulate to a new key entirely.
The samples go at random as well:
- Which animal sound we play.
- How fast it plays (down to half speed, up to double).
- Whether it plays forward or backward.
- How long to wait until playing another one.
This was my first non-trivial project with Sonic Pi, and I'm sure I footgunned hard here, but I had a ton of fun with it.
Fake Animal Facts
This was built much like the Personal Climate Impact sound effects: I pulled a generic background soundscape from YouTube, and layered my voice recordings over that soundscape.
Thanks for the read! Disagreed? Violent agreement!? Feel free to drop me a line at , or leave a comment below! I'd love to hear from you 😄