AI PCs are here, and they’re bringing new possibilities. Audacity has introduced on-chip audio AI tools for music generation, transcription, and more.
Intel showcased these tools in Malaysia, using Audacity as a demo partner while describing the Meteor Lake (now Core Ultra) architecture. The released tools utilize OpenVINO, an open-source toolkit developed by Intel that has been optimized for Intel’s 11th-gen Core chips and Arm architecture, but not AMD or Ryzen AI.
The new AI tools from Audacity include:
Noise suppression, which filters out background noise from music and conversation
Transcription, powered by Whisper.cpp, which can transcribe words to a label track and an external file
Music generation, which uses Stable Diffusion and Riffusion to generate music from a prompt or existing music
Music separation, which can split a song into separate instruments, including vocals
However, these new AI tools have some limitations. They only work with an older version of Audacity, specifically Audacity 3.4.2. Audacity has provided instructions for installing the new tools, though, which are relatively straightforward.
To install the new tools, you’ll need to download Audacity 3.4.2 as well as two files: Audacity-OpenVINO-Module-3.4.2-R1.zip and openvino-models.zip. The models file is 1.92GB, and once downloaded, you’ll need to copy the files to the Audacity installation folder. Then, run Audacity.exe, go to Edit>Preferences and change mod-openvino to “Enabled”. Close and reopen Audacity to activate the new tools.
With the installation completed, you’ll have access to these early AI PC tools. Enjoy!