Every sound you hear in this song was made from human voices.
I wandered around with a little microphone at GDC San Francisco 2011 to record vocal sounds from anyone who was willing to make them. I wanted to make a song out of these noises. In fact, I wanted to make a complete song, using nothing but these vocal recordings.
This is that song.
People grunted, groaned, moaned, screeched, screamed, and sang into my little microphone and I used this source material to produce Voices of Devs.
Absolutely everything you hear in this track was made by manipulating these recordings using various sound design techniques. Every drum, lead, chord and effect was, at one point in time, a recording of somebody making noises with their mouth at GDC San Francisco 2011. No other samples, synths, instruments or recordings were used in this track.Whatsoever.
How was this achieved?
The goal was to manipulate these original vocal recordings using various musical sound design techniques and turn them into various musical instruments – drums, synths, etc.
The entire process boiled down to two techniques:
1. Matching waveform visuals
2. Looping waveforms for “synths”
For those of you playing at home, here’s a breakdown of each process!
Matching Waveforms Visually
The first technique is very simple, and it’s probably closer to a visual artist technique rather than something musical. The idea is that you visually inspect the waveform of a sound that you’re trying to replicate, and then attempt to match it using whatever source material you have (in my case, the recordings from GDC).
Take this image on the right – this is the waveform of the stock Kick Drum sample from FL Studio 10. Click below the image to hear it.
Visually we can see that the kick is basically made up of two elements:
1. A crazy looking transient at the start.
2. A sine wave that goes up and down about 12 times whilst being stretched.
Now, all we need to do is create something that looks visually similar (we’re not worried about sound at this stage) using our source material.
I used these three elements to create the “transient”:
Next, I built a sine wave from a single sample:
The transient samples were sliced up, pitched and EQd, and then combined with the sine sample using the original kick sample as a visual reference. The final kick drum sample was then EQd, compressed and tweaked until I ended up with this:
This technique of visually reproducing the sound I wanted to create was used to create all the kicks, snares and cymbals you hear in the track.
Waveform Looping and Resampling
The second technique is also very simple and is based on Wavetable Synthesis, where a segment of digital audio is looped to create a constant tone that can then be manipulated via a sequencer. This technique is extremely simple and allows for a lot of freedom and experimentation with virtually endless possibilities.
Step 1 - Grab A Sample
The fun part of this technique is the experimentation. First off, find any sample of audio – anything at all. For this, I’ve grabbed a cool screech that Steve Tushar gave me.
Step 2 - Snip A Loop
Explore the different parts of the sample by zooming-in and looping little chunks. Listen for anything cool or unique sounding because this will become our waveform that we will use for synthesis. I like this little section of Steve’s screech because it’s a bit aggressive and angry and kinda “gnars”, and I’m feeling kinda gnarly today. Once you’ve settled on your little chunk, copy it, paste it into a new file, fade the edges so it loops on zero crossings and save it as a wave file.
Step 3 - Pitch It
Next we need to (well, I guess it’s optional) pitch correct the waveform chunk so it conforms to concert pitch. This will make sure it’s in tune with everything else and make it easier to sequence. This waveform was pretty close to an “E” note.
Step 4 - Load Into Sampler
Next we load the little waveform chunk into our favorite sampler, make sure it’s looping and stretch it out over the keyboard. This will automatically pitch shift the sample based on the key you press, and we’re now beginning to see our waveform synth take shape!
Step 5 - FX!
Next we want to shape and manipulate the waveform in the sampler to give it some more character. Here I’ve added a Tape Saturator for distortion and grit, I’ve EQd it to make it brighter, added a chorus and phaser, compressor and then some delay and reverb. I’ve also added a Formant filter which is being modulated by an envelope sequencer to give it a slight “talking” effect which is popular among all the young synth-groovers of today.
Step 6 - Program The Sequencer
Programming allows a further level of control which further expands our possibilities. To give our waveform synth a real “synthy” touch I’ve given it a portamento script to make it jump between notes. After some more buss effects, EQ and compression we get this!
All this from one little screech vocal sample. This is a very basic demonstration and this technique can be vastly expanded by layering multiple waveforms, more effects, programming, etc. All synth, bass, plucks and guitar-type sounds you hear in the track were created with this technique.
And there you have it!
A lot of work went into this track, a lot more than a single post can handle, but these two main techniques formed the majority of the sounds that you hear in Voices of Devs. If you have any questions please feel free to email me, find me on Twitter, or Facebook.
This was so much fun to put together and I’d like to thank all the awesome people who contributed their voice to my little project: Jeff Ball, Sean Beeson, Alex Brandon, Nathaniel Chambers, Ben “Crossbones” Cooper, Charles Deenen, Brian DiDomencio, Morla Gorrondona, Joe Griffith, Lance Hayes, Kole Hicks, Jimmy Hinson, Thomas Kohl, Sam Hulick, Jake Kaufman, Kieran Lord, Kunal Majmudar, John Mazzei, Hazel Mckendrick, Ben Minto, Keith Moore, Lennie Moore, Jayson Napolitano, James Nixon, Dan Reynolds, Pontus Rufelt, Michael Schiciano, Paul Sivertsen, Jeremy Taylor, Hillary Thomas, Joe Thwaits, Steve Tushar and Mike Worth.
Update - 4 years on
Things evolve so quickly these days and looking back on this project I see I'm still using these techniques frequently, however the ease in which these techniques has gotten much easier. Ableton Live's sampler and various other built-in tools would make a project like this much easier to tackle!
Here you can download the Kontakt patches:
If you like the track you can purchase it here:
Here are the multitracks for further study: