Adobe Reveals New AI Music Maker Tool - Generates music from text

Adobe is dipping its toes into musical AI with a new experimental program called Project Music GenAI Control. The tool allows users to generate original music and audio directly from text-based prompts. 

Putting words to song, the program can turn phrases like “powerful rock” or “sad jazz” into custom instrumental pieces. It draws from genres to craft intros, outros, and background tracks tailored for any project. 

From those initial compositions, creators get hands-on control. Within the interface, they can refine tempo, structure, repetition and intensity to perfectly fit their needs. Editing options let producers transform generated sections or create loops to taste. 

In addition to flexibility, Adobe promises the AI assistant is a safe creative partner. Any resulting music can be commercially released. Moreover, the company plans to integrate its “nutrition labels” system for transparency. Labels will disclose the level of AI involvement in any given work.

This one-stop audio shop stems from a collaboration between Adobe Research, University of California San Diego, and Carnegie Mellon University. It brings music making and editing together in a single workspace to streamline production workflow pains.

While still in testing, the tool could shape how creatives craft soundtracks going forward. Only time will tell if Adobe incorporates it into their popular Creative Cloud suite down the line. For now, the experimental nature means a consumer version may be a ways off.

Stay tuned for updates on if and when Adobe’s music making machine hits the mainstream. In the meantime, behind-the-scenes tinkerers get an early look at where machine learning is headed next in media creation.