audio-reactive glsl sandbox forked from doob's glsl-sandbox

for rapid prototyping of concert visuals

NOTE: my additions were written under hackathon-style conditions, so please excuse the mess (for now)

setup

1. bundle install
2. export MONGOHQ_URL="mongodb://[user]:[password]@127.0.0.1:27017/[database]"

run it

1. start mongo (mongod)
2. bundle exec rackup config.ru
3. hit http://localhost:9292
4. (optional) generate fake certs and use thin for SSL so you don't get a prompt for microphone access every time you visit the shader page: thin start -p 9292 --ssl --ssl-verify --ssl-key-file .ssl/server.key --ssl-cert-file .ssl/server.crt

shader uniforms

so you don't have to deal with GLSL's const-only array indexing, there's a data-to-texture pipeline included. declare uniform sampler2D audioTex in the frag shader, then access the data anywhere in the shader with vec4 datum = texture2D(audioTex, gl_FragCoord.xy);. you can replace gl_FragCoord.xy with any vec2 with normalized coordinates.

todo

1. controls for fading/transitioning
2. should rewrite this with a less bulky interface & less overhead ("presentation mode"?)
3. audio smoothing
4. a lot more