Open gyuchulm opened 1 year ago
(
// Server.default.options.inDevice_("BlackHole 16ch");
// Server.default.options.outDevice_("BlackHole 16ch");
Server.default.options.inDevice_("MacBook Pro Microphone");
Server.default.options.outDevice_("MacBook Pro Speakers");
// Server.supernova;
Server.scsynth;
s.options.sampleRate = 44100; // the sample rate must match RAVE
s.options.blockSize = 512;//16; // a low block size within scsynth is fine,
s.options.hardwareBufferSize = 2048
; // but the hardware buffer size needs to match RAVE
// (to avoid dropouts)
s.waitForBoot{
~bus = Bus.new(index:0, numChannels:2);
}
)
(
b = Buffer.read(s, Platform.resourceDir +/+ "sounds/a11wlk01.wav");
~synth = {
var z = RAVEEncoder.new(
"/Users/gyuchulm/Downloads/percussion.ts", 4, PlayBuf.ar(1, b, BufRateScale.kr(b), loop: 1.0),// input for latent embedding
);
// z[1] = MouseX.kr(-3, 3);
// z[4] = MouseY.kr(-3, 3);
Limiter.ar(
RAVEDecoder.new(
"/Users/gyuchulm/Downloads/percussion.ts",
z //latent input
)
)!2;
}.play;
)
s.reboot;
s.quit;
s.boot;
(
~synth = {
var z = RAVEEncoder.new(
"/Users/gyuchulm/Downloads/organ-v2-test_streaming.ts", 8,
Dust.ar(4, 1),// input for latent embedding
);
// z[1] = MouseX.kr(-3, 3);
// z[4] = MouseY.kr(-3, 3);
Limiter.ar(
RAVEDecoder.new(
"/Users/gyuchulm/Downloads/organ-v2-test_streaming.ts",
z //latent input
)
)!2;
}.play;
)
still can't load model with trained V.2
Hi, I built the new version, but still have some issues with same before. (only works pre trained model with v1)
I use Mac and I built in sc in user (root) folder, and built an extensions in root and move Rave class in SC in Library-> Application support. should I need to do something more?
and where is the proper location of plugins/RAVE? should be in an Applicaition->Supercollider-> package contents?
But still have some issues with playing v2 model.