© EPA/Hannibal Hanschke
Stable Diffusion is a well-known name in the world of generative AI. The company behind it, Stability AI, now has one New software releasedWhich specializes in composing music. Stable sound It is the name of an artificial intelligence tool Entire pieces of music from just a few text entries creature.
Specifications can be “Trance, Ibiza, Beach, Sun, 4 AM, Progressive, Synthesizer, 909, Dramatic Chords, Choir, Euphoric, Nostalgic, Dynamic, Flowing”. Text written by Amnesty International Only in English Understood.
Trains on 800,000 addresses
As reported by The Verge, Stability AI claims to have trained its new software on 800,000 audio files. In general, the data set should be: 19,500 hours of recordings Composed. also Title with copyright protection there. Stability AI has the right to use this material through a license agreement.
➤ Read more: Grammy Awards for AI music: New rules specify details
Try it for free
Those interested can use static audio Free Uses. Maximum 20 tracks per month and max 45 second tracks Are the restrictions. In addition, there is another Professional levelWhich can be used to create 500 tracks per month with a length of up to 90 seconds. You pay $11.99 per month. Of course, 90 seconds is still not enough for a commercial music clip of the usual length.
Creating music from text is Nothing entirely new. Meta, for example, offers an AI called AudioCraft. Google owns MusicLM. However, both programs can currently only be used for specific test subjects.
