iO-808 VS Magenta Studio

I would highly recommend enabling click and drag to "paint" notes. As it is right now, if I want 16 closed hats, I have to move, click, 16 times. I'd rather drag to paint based on whatever state of the note I start on. The mutes on the left would be better if they mute the notes, not the sounds. Muting and then enabling can end up playing the tail of some of the longer sounds. This isn't typically how you want... - Source: Hacker News / about 1 year ago

Here is a classic 808 drum machine for your browser. Have fun! :-). Source: almost 2 years ago

If that’s just about making beats, a Roland TR8 is awesome for that. Or even a Teenage Engineering PO Rythm, a jumbee or make https://io808.com his homepage. Source: over 2 years ago

In fact, try it right now. Go to https://io808.com/ . Set the tempo dial to 3. For the bass drum (BD), enable steps 1, 7 and 8. For the snare drum (SD), enable steps 9, 15 and 16. Source: almost 3 years ago

Also of note that Firefox's Web Audio API implementation just isn't very good in general. It's my daily driver, but I won't run Airsonic in it, because after half an hour or so the music reliably starts glitching. Fine in every other client, so it's definitely a Firefox thing, and iO-808 [1] also calls it out in an alert if you go there in Firefox. Granted, a glitchy audio implementation might be just the thing... - Source: Hacker News / about 3 years ago

Check out Magenta on Tensorflow https://magenta.tensorflow.org. It can generate, interpolate or continue midi files, which can be inserted into Ableton. Source: 11 months ago

Yes, most models these days, except the exceptionally large ones, are possible to train on a laptop. Of course it helps if your laptop has Nvidia CUDA GPU, but even if it doesn't you can rent an AWS 4 core/16GB GPU instance for 0.5 cents an hour. 24 hours of training time would be quite a lot for most models, unless you're trying to train a FB any to any language type model, but typically the big huge models are... Source: over 1 year ago

Sounds like you're referring to this https://openai.com/blog/jukebox/ or this https://magenta.tensorflow.org/. Source: over 1 year ago

Google Magenta has also put out interesting neural net experiments with music like a VST which can blend sounds together to become an instrument, and a Google Colab notebook which attempts to make a MIDI from an audio file. Source: over 1 year ago

Magenta is an open-source Python package built on top of TensorFlow to manipulate image and music data to train a machine learning model with the generative model as the output. To learn more about Magenta here you go. - Source: dev.to / almost 2 years ago