I built a review-first tool for reducing steady background noise in audio and video
I built NoiseVanish for a narrow problem: reducing steady background noise around spoken-word audio and video while making the result easy to review before someone relies on it.
Instead of treating a quieter waveform as proof, the workflow keeps the original and cleaned versions available for a same-position comparison. The main target is steady noise such as fan or air-conditioning wash and electrical hum. It does not promise to remove every transient sound, and results vary with the source.
I would value feedback on two things: whether the before/after review flow makes the limitations clear, and where people still hesitate before processing a real recording.
The before/after review is solving a measurement precision problem, not just a UX problem. Users hesitate because they can't measure what the tool will do to their audio - they get a binary choice (process/don't process) with no visibility into artifacts or edge cases. Your measurement system (side-by-side comparison) creates a signal that a verbal description or spec sheet cannot: "here is exactly what this choice produces." The hesitation doesn't disappear because the review shows risk - it becomes informed hesitation. That's the win. The question is whether you measure what users actually measure when they decide - do they listen for specific artifacts, trust frequency content, check speech intelligibility? If you track what they compare most, you've found what "good" actually means to your buyers.
My hesitation before uploading a customer call would be data handling, separate from audio quality. I'd want the upload screen to explain where processing happens, recording retention, and deletion. A built-in sample would let me try the review flow before uploading anything private.
The review-first workflow seems like the real differentiation. Have you seen users who initially hesitate actually become comfortable processing a recording after comparing the two versions, or does the hesitation mostly remain because they still don’t know what artifacts to expect?