aap-core VS NeuralNote

I know Atsushi Eno has been working on this for quite some time now: https://github.com/atsushieno/android-audio-plugin-framework

Scalar 2 is beneficial for sure, but definitely is best when accompanied with music theory knowledge. I think it's great for when you want to try some different things, but don't want to figure all the potential variations of some chord in F#m, because no matter how much theory you know, that'll always take at least some brain power. A lot of the other ones ive seen seem to overpromise and are essentially just random number generators. Things like abletons scale device are handy. I can't play all the keys on piano as easily as I can play Cmaj/Am so using that to be able to play white keys, but match the song's key is helpful. And if you ever want audio to midi to get the midi of a sample of something, I think BasicPitch is the best algorithm. Better than ableton or melodyne in my opinion. This free VST uses it under the hood. It also lets you adjust sensitivity, min. note length, and the threshold for splitting a note, which can give you less of the "drunk with the shakes" vibe that audio to midi usually has. That being said, it's not always better, there have been times that Ableton or Melodyne does better for a given sample, but more often than not Basic Pitch does the better job.

There's this https://basicpitch.spotify.com/ and this https://github.com/DamRsn/NeuralNote which work well and are free, but aren't perfect. There's others too (search for polyphonic pitch to midi converter)