Skip to content

Audacity Is Now Even Smarter with New AI Plugins for Music & Podcasts

Photo by Caio Silva / Unsplash

Intel has unveiled a set of AI-powered plugins for audio editing, music generation, and podcasting within the popular Audacity software. These tools, built on the Intel OpenVINO AI toolkit, aim to enhance the precision of audio editing processes. In total, six plugins have been integrated into Audacity, each leveraging well-established artificial intelligence models.

The Music Generation and Music Style plugins utilise the Stable Diffusion model and its Riffusion branch. These tools generate music directly within the program, either based on user-provided textual prompts or existing audio tracks.

Music Generation focuses on creating background tracks, while Music Style is designed to alter the style of existing compositions. Additionally, Audacity now can separate audio content from finished compositions. With the Music Separation plugin, users can isolate vocals from accompaniments or break down entire songs into individual tracks featuring bass, drums, vocals, and other instruments. As an extension to this plugin, the tool has also acquired the Stem Separation tool, which further divides compositions into stems.

Image credit: Audacity

Incorporated for the editing of recorded signals, the "smart" Noise Suppression and Whisper Transcription tools have been added. The former eliminates any background noise from recorded signals, while the latter transcribes whispers and regular speech into text. According to developers, both tools will be particularly beneficial for podcast creators.

Martin Keary, Audacity's Product Manager, stated that the addition of AI-powered plugins aims to make Audacity more attractive to both new and existing users. He emphasised the intention to "offer capabilities that were previously unavailable or simply difficult to achieve." Speaking with the Intel Business YouTube channel, Keary clarified that "the plugins work locally on the user's computer, not on a remote server, which is driven by privacy concerns." Keary referred to the integration of intelligent sound processing and generation tools as "just the first step" for the Audacity team, indicating plans to expand the library of similar plugins in the future.

The smart plugins will not be bundled with Audacity; instead, their installation packages have been made available in the official repository on the GitHub platform. It's noted that due to several reasons, the functionality of these tools is currently limited to the Windows environment. The timeline for the release of versions of the tool for macOS and Linux, where these smart assistants will be operational, remains unspecified.