Meta AudioCraft needs text-based prompts to create music. (Image: Meta)
Ever thought of a ChatGPT-like tool, but for generating music? Meta has done exactly that with its AudioCraft generative AI tool. Read on to know how it works.
Creating original soundtracks for projects of all kinds is a time-consuming and expensive process that requires the expertise and skills of an artist. But what if an AI-powered tool could create a custom soundtrack suited to your project with just a few prompts? Meta has done exactly that with its AudioCraft generative AI tool. Meta promises that users can create “high-quality, realistic audio and music from text” with the tool family.
Meta notes that while AI generated images have taken off—AI-generated audio is still lagging behind, but with the new AudioCraft tool family, users will be able to create “music, sound, compression, and generation—all in the same place.”
The tool is divided into three models: AudioGen, MusicGen, and EnCodec. AudioGen generates audio from text prompts based on public sound effects, while MusicGen does the same thing but with music licensed by Meta. The EnCodec decoder—on the other hand—allows for higher quality music generation with fewer artifacts.
Users Can Generate a Slew of Sound Effects
Meta also claims to be releasing pre-trained AudioGen models, which will let users create environmental sounds and sound effects—such as a dog barking, cars honking, or footsteps on a wooden floor.
“The AudioCraft family of models are capable of producing high-quality audio with long-term consistency, and they’re easy to use,” Meta said.
“With AudioCraft, we simplify the overall design of generative models for audio compared to prior work in the field—giving people the full recipe to play with the existing models that Meta has been developing over the past several years while also empowering them to push the limits and develop their own models,” it added.
Meta Has Made AudioCraft Tool Family Models Open Source
Meta is also making its audio generation tools open source, like its Llama 2 LLM, so that researchers and developers can create their own models. Meta claims that by open sourcing the models, it can help advance the field of AI-generated audio.
