Since the workflow for messing with these is always the same I decided to make a little shell script to speed up the process. I’m on a Mac so I have to use sox for playback. If you are on Linux you’d need to change the pipe to /dev/audio or /dev/dsp unless you wanted to use sox instead.


#!/bin/bash

# This script creates and plays a simple ByteBeat
# ARGS
# $1: a string with the ByteBeat algorithm e.g. "((t * 3) & (t >> 5))"
# $2: the name of the file to be creates (without an extension)
###
# make the c file
echo "main(t) {
for( t = 0;;t++)
putchar( $1 );
}" > "$2.c"
###
# compile the source
gcc "$2.c" -o "$2"
###
# play it with standard ByteBeat settings
./"$2" | sox -traw -r8000 -b8 -u - -tcoreaudio


Save that to a file called something like bytebeat.sh and make it executable with chmod 755. after that you can just run

./bytebeat.sh "((t * 3) & (t >> 5))" test

and automagically you get 8-bit goodness right from your terminal. There’s a slight delay between running the script and hearing the output because it needs to compile so I wouldn’t necessarily perform with this (unless you get really good a timing it). Now this could be modified to let you specify the arguments for sox or whatever… but meh. This is just intended as a quick and dirty way to look like you are more of a hacker than you really are.

Homunculus from Cole Ingraham on Vimeo.


#!/bin/bash
for i in {1..13}
do
say "a night with Phillip Glass followed by"
done
for i in {1..11}
do
say "a night with Phillip Glass followed by"
say "a night with Phillip followed by"
done
for i in {1..8}
do
say "a night with Phillip Glass followed by"
say "a night with Phillip followed by"
say "a night with followed by"
done
for i in {1..5}
do
say "a night with Phillip Glass followed by"
say "a night with Phillip followed by"
say "a night with followed by"
say "a night followed by"
done
for i in {1..3}
do
say "a night with Phillip Glass followed by"
say "a night with Phillip followed by"
say "a night with followed by"
say "a night followed by"
say "a followed by"
done
for i in {1..2}
do
say "a night with Phillip Glass followed by"
say "a night with Phillip followed by"
say "a night with followed by"
say "a night followed by"
say "a followed by"
say "followed by"
done
say "a night with Phillip Glass followed by"
say "a night with Phillip followed by"
say "a night with followed by"
say "a night followed by"
say "a followed by"
say "followed by"
say "by"


Canvas from Cole Ingraham on Vimeo.

SuperCollider has a method called ratiomidi that takes a ratio and converts it to the float MIDI number version. This replaces the need to multiply the ratios by a fundamental and mod 12 it as, because all our ratios are within an octave, everything will be from 0.0 through 11.99. In other words: it gives us tuned versions of pitch classes. This one simple method removes 2 of the previous steps.

[1,17/16,9/8,6/5,5/4,4/3,10/7,3/2,8/5,12/7,7/4,15/8].ratiomidi;

gives us

[ 0, 1.0495540950041, 2.0391000173077, 3.1564128700055, 3.8631371386483, 4.9804499913461, 6.1748780739571, 7.0195500086539, 8.1368628613517, 9.3312909439626, 9.6882590646912, 10.882687147302 ]

Now that we have that, like before we need to get the cents in the +/- 50 cents, here in decimals. Once again we just need to subtract a series of integers from 0 to 11 like so:

( [1,17/16,9/8,6/5,5/4,4/3,10/7,3/2,8/5,12/7,7/4,15/8].ratiomidi – (0..11) ).round(0.01);

which gives us

[ 0, 0.05, 0.04, 0.16, -0.14, -0.02, 0.17, 0.02, 0.14, 0.33, -0.31, -0.12 ]

Great, that was easy! Now, you may notice by comparing it to the result we got from the other post that although the numbers are all the same, they aren’t in the same order. That is because before we based the tuning on A being the fundamental/tonic. Here C (array slot or pitch class 0) is the tonic. This is fine if you want C to be the tonic but what if you don’t? Simple! Arrays in SuperCollider also provide a method called rotate(numPlaces) which, as you might guess, shifts items left or right by a number of slots. So for our final version of this:

( [1,17/16,9/8,6/5,5/4,4/3,10/7,3/2,8/5,12/7,7/4,15/8].ratiomidi – (0..11) ).round(0.01).rotate(9);

gives us

[ 0.16, -0.14, -0.02, 0.17, 0.02, 0.14, 0.33, -0.31, -0.12, 0, 0.05, 0.04 ]

which is exactly what we got in the old example. Why are we using 9? If you prefer to think about this as a transposition operation then T9 = A. You could also put a -3 in there and you would get the exact same result. Totally depends on how you want to think about it.