On Monday, Nvidia unveiled Fugatto, an AI tool that promises to create sounds never heard before. This groundbreaking technology allows users to generate unique music, sounds, and speech from text and audio inputs it has never encountered.
Imagine a trumpet that meows or a saxophone howling into electronic music with barking dogs. Fugatto takes wild ideas and turns them into reality by offering endless possibilities for sound creation.
This groundbreaking tool allows users to give prompts that can produce completely unique audio effects. For example, you can ask it to create “deep, rumbling bass pulses paired with high-pitched digital chirps, like a massive sentient machine waking up.” Fugatto takes these detailed descriptions and transforms them into sounds that no one has ever heard.
Fugatto stands apart from other AI audio tools, such as those from Stability AI, OpenAI, and Google DeepMind, by creating entirely new and unheard sounds. While AI-generated music and audio tools are becoming more common, Fugatto’s ability to invent new sounds makes it a game-changer.
To create Fugatto, Nvidia’s researchers gathered millions of audio samples and developed a new set of instructions to expand the tool’s capabilities. This allows the model to handle more tasks with greater accuracy, all without needing additional data.
The technology can also edit existing music, such as isolating vocals, adding instruments, or even replacing a piano with an opera singer.
Nvidia has not yet shared when Fugatto will be available to the public, but the tool’s potential is clear. As AI audio tools continue to evolve, Fugatto could be at the forefront of creating a new world of sound, offering musicians, sound designers, and creators endless possibilities.
With Fugatto, the future of sound could be as limitless as your imagination.
