Adobe launches audio processing tool to enhance speech quality

Adobe has released an AI-based audio processing tool to enhance voice speech. Called Enhance Speech, this software removes background noise and makes the voice sound clearer. 

At present, Enhance Speech can be used for free by those with an Adobe account. Users can upload an MP3 or WAV file of up to 1 GB in size or one-hour long duration.  It could be accessed from the browser directly and the output enhanced audio snippet can be downloaded.

On discussion forums like ycombinator, users largely gave positive reviews of the tool, saying that it helps in producing studio-like quality. That said, a few users complained that the resulting snippets might be understandable but sound ‘unpleasant’ and in some cases, different from the original speaker. 

This software is part of the Adobe Podcast project that was launched in 2021 as Adobe Shasta. It started as an Adobe Labs experiment to find ways to improve the way users edit audio on the web. Writing about the project, Adobe’s head of audio products said that pandemic made recording difficult for even audio professionals and so the initiative was born to assist in creating professional sounding audio. Under Project Shasta, Adobe has introduced several innovations like remote recording, mic check, text-based audio editing, and AI filters. Recently, Project Shasta was rebranded to Adobe Podcast.