Adobe
Adobe recently released a free AI audio processing tool that can improve some poor-quality voice recordings by removing background noise and making the voice sound stronger. When it works, the result sounds like a recording made in a professional sound booth with a high-quality microphone.
The new tool, called Enhance Speech, was created as part of an AI research project called Project Shasta. Recently, Adobe Project has rebranded Shasta as Adobe Podcast.
Using Enhance Speech is free, but requires an Adobe account and works best with a desktop web browser. Once registered, users can upload an MP3 or WAV file of up to one hour or 1 GB. After a few minutes, you can listen to the result in your browser or download the resulting cleaned audio.
In our tests with the service, Enhance Speech worked best with audio that contained a voice without crosstalk or excessive noise. For example, we recorded audio from the built-in microphone of an iMac of a person standing 10 feet away, including the sound of a nearby fan, and the resulting audio (after processing by Enhance Speech) sounded as if it had been recorded up close in a noise-free environment. studio with a professional microphone.
Adobe
How does it work? Adobe didn’t provide details, but we suspect the company has trained a deep learning model on many (possibly thousands) of hours of clean and noisy audio. The model could then “learn” to select the human voice frequencies and synthesize a facsimile that accurately matches the source. This is speculation until Adobe provides more technical details and we’ve contacted the company for comment.
On that note, some Hacker News commentators have reported hallucinatory results – unexpected output such as phantom voices where the AI misinterprets the input audio – from extremely noisy audio (such as speech recorded next to a waterfall) or from non-English sources, suggesting that Enhance Speech does more than just a conventional noise reduction technique.
Enhance Speech isn’t the first tool to offer this kind of AI-powered noise reduction capability. For example, an open source package called mayavoz and a commercial service called Audo Studio do something similar.
It’s worth noting that Enhance Speech is part of a larger group of AI-powered podcasting tools from Adobe, including a Mic Check tool (also currently available for free) and a transcript-based audio editing tool that’s still in beta testing by invitation only.
