Moises AI: The Musician's App

Seriously, that is way annoying! If you find bugs daily, then it is alarming!

This is freaking amazing and so easy to use

Indispensable pour musicien en pratique solo

Amazing quality

This app is amazing 👌🏼

Separates parts of songs extremely well, I am extremely impressed

There isn’t much out there that does what this does and it does it well enough. My only complaint so far is the difficulty the app has splitting harmonies but it’s totally understandable and easy enough to work around. If you’re a vocalist or instrumentalist looking to make some practice tracks, or just want to pitch shift your own karaoke tracks this works more than well enough.

They should add a feature to select parts of the song and train the AI what the actual lyric is per part of the song. The AI generation gets lost in tempo when there is any chorus in the song to it details the song and it’s so generated lyrics. There should be an option to regenerate it, instead of manually edit the lyrics because when you set timing and lyrics manually on the chrome browser, or on desktop the lyrics when you set the exact time on the timeline move around due to some non snapping clip bug where they side around due to playback. And there should be a feature where it splits instruments in separate depths because most songs you try to split and listen to have different mastering styles where the songs I wanna listen to are from the 80’s and the synth track is so hard to get with this app even with pro because this splitting algorithm doesn’t apply or have the depth layers in mind. This would be a good feature if the split algorithm also had a (speaker depth) gui or graphics where you can choose (Far Left) (Left Middle) (Middle) (Right Middle) (Far Right) Or an additional depth layers for specific echo effects. And may recreate the echos by isolating the Dry echos and matching the delays. Or removed the mid vocals and combine the delay or echo of the vocals, as one separate layers to be the (echo layer: vocals). And the Stereo Dry echos can be one layer. (Vocals) >(echo layer) (Backing vocals) >echo Generation example (Vocals) (If echo is decided it does a deep dive learning and attributes the echos again a database of known echo and it effects and calculates the phase and timing/delay differences to determine the echo and what echo effect it uses.) (Chooses the likely echo) (Ping-Pong 0.00mls(etc) vocals) So like, (Vocals) (Backing vocals) And like this when open is enabled: (Vocals) |_(Ping-Pong ‘+Stereo Delay’ +00.0mls -00.0mls Vocals) (Backing vocals) |_ (Stereo Spread +00.0mls -00.0mls vocals) Also this algorithm should be possible to get synths better out of any song because most synths from the 80s have synths in the depth field instead of the middle of the song. If you’re reading this on mobile: The example is more like this: (Vocals) |_(Ping-Pong 00.0mls vocals) (Backing vocals) |_(Stereo Spreader 00.0mls BV) This may take longer to separate but it would be at a much better expense than what we have right now. For songs with depth that this algorithm cannot separate due to layers that the separator cannot hear d

Très bonne application. Le seul point faible que j’ai trouvé c’est que cette application est payante mensuellement et non à un prix à l’achat comme plusieurs application musicale que l’on paie une fois et c’est tout.

Logic for iPad subscription is much cheaper/ more powerful stem separation. Apple came to our rescue with a historic port of Logic pro to iOS. M’s price should have went down not up or not been an uncompetitive subscription.