Just in case you scrolled past it, the live demo was in the github website link:<p><a href="https://idroppedmyphonethescreencracked.tumblr.com/" rel="nofollow">https://idroppedmyphonethescreencracked.tumblr.com/</a>
WebAudio-based library that provides quick way to set up a web audio graph.<p>Desktop wrapper:
<a href="https://github.com/billorcutt/Cracked">https://github.com/billorcutt/Cracked</a><p>Cat examples:
<a href="https://idroppedmyphonethescreencracked.tumblr.com/" rel="nofollow">https://idroppedmyphonethescreencracked.tumblr.com/</a>
Can it handle "nodes" that emit a different number of audio samples than they consume?<p>I'm thinking of time stretch effects like mine <a href="https://github.com/bungee-audio-stretch/bungee">https://github.com/bungee-audio-stretch/bungee</a>
if you are looking for some performant declarative web audio lib in js, check:<p><a href="https://glicol.js.org/" rel="nofollow">https://glicol.js.org/</a><p>it's ported from Rust
I find the underlying premise a bit odd. I can name values in Javascript just fine:<p>const whatever = ...<p>I would rather refer to them by these names than by strings. It's both faster and safer to do so.