Despite only being released in April of this year, Uncharted Labs' AI music creator Udio continues to get more robust. Version 1.5 has just been released and introduces many audio improvements along with several new features.

According to Udio, the generator now creates 48 kHz stereo tracks with enhanced clarity and consistency. The announcement claims that “clarity, instrument separation, transients, coherence, and musicality are improved across generations.”

In the example provided by Udio, there was an improvement in sound. In version 1, the lyrics sounded more like tin cans; in v1.5, they sounded warmer and more authentic.

The other noticeable improvement is that the audio sounds like it is actually in the speakers. In the example provided, the audio mix sounds like it was recorded from across the room and sounds echo-like; the 1.5 version feels like it was mixed next to the microphone and, for lack of a better word, sounds closer to the speakers.

Other new features include a new creator page, stem downloads, audio to audio, and critical controls.

Stem download allows users to split a fully mixed track into four stems (vocal, bass, drums, and “everything else”). You can take out parts of the song for remixing, or use external tools to modify the song.

The key control is interesting because it can prompt the song to be in a specific key, such as C minor or A major. However, Udio warns that this feature is still a work in progress and the result may not hit the intended key at all.

Finally, version 1.5 improves global language support and adds new languages such as Mandarin. However, Udio does not specify how many new languages were added in this update. u

Those interested in testing Udio should check out the guide before getting started.