ElevenLabs, a company specializing in AI audio, is expanding the capabilities of its AI music generation service, 'ElevenMusic,' to strengthen the personalized music production environment.

Eleven Labs announced on the 23rd that it has released new features, 'Vocals' and 'References', through a major update to Eleven Music. In addition, the scope of support for the 'Styles' feature has been expanded based on the latest generation model, 'Music v2'.

The core of this update is to provide a more consistent AI music generation environment by leveraging the user's voice and musical style.

The newly added 'Reference' feature is a function that generates new music by reflecting the mood and style of a song when a user uploads an audio track ranging from 10 seconds to 5 minutes in length.

Eleven Labs explained that it can convey musical nuances that are difficult to express with text prompts alone to AI models through actual audio sources.

They also stated that specific sections can be reworked in conjunction with the Inpainting function, which allows editing only parts of the music, and that the system is designed to ensure all uploaded audio undergoes a copyright check.

The 'Vocal' feature is a function that creates new songs using the user's voice or vocal library.

Users can upload voice recording files to create songs that preserve their voice, lyrics, and timbre characteristics. It also supports the creation of reusable vocal profiles by learning vocal characteristics from existing music files.

It is also possible to create songs using vocals provided in the library without separate voice registration.

The existing 'Style' feature has officially supported the Music v2 model. It supports maintaining the overall sound characteristics of the music by utilizing user-created styles or existing style libraries.

Eleven Labs explained that combining vocal and style functions can create a music production environment where voice and sound are consistent.

The company stated that it maintained copyright protection as a key principle in this update as well, and announced that it performs copyright checks on uploaded voice and sound sources at the pre-creation stage.

The new features are immediately available through the ElevenMusic platform, and some features are also provided via ElevenCreative and API services.