I was skeptical about Studio Sound being anything more than a basic noise gate. Had to test it myself.
I recorded an interview on a busy downtown sidewalk. The raw audio was a mess:
* Constant traffic rumble
* People talking in the background
* Wind hitting the mic
Ran it through Studio Sound (default settings). The difference is stark.
You can hear the samples here: [link to audio samples]
Key takeaways:
* It removed the steady traffic noise almost completely.
* Voices are clearer, but there's a slight "underwater" effect on the sibilance.
* It's not studio quality, but it's usable for a quick social clip. Wouldn't use it for a full podcast without some post-processing after.
For a free feature, it's impressive. It saved me an hour of manual cleanup in Audition. ROI on my Descript subscription just went up a notch.
Just the facts.
Trust but verify.
That "underwater" effect on sibilance you noticed is a classic trade-off with these AI tools. They're fantastic for removing broad-spectrum noise like traffic, but they often chew up the high-end frequencies a bit.
You're spot on about the use case - perfect for a quick social clip where speed matters most. For anything longer form, like your podcast example, I'd still want to take that cleaned output and run a gentle EQ pass to try and bring back some of that lost clarity.
The hour saved in Audition is the real win here. Time is the one resource you can't get back.
ship early, test often