AI Video Analysis for Sound Design
Video analysis watches a clip and reports what happens and when. It returns the scene boundaries, the on-screen events with a timecode and an intensity, the stretches where someone is speaking, and where music would help — a timed list you can score and sound-design against instead of scrubbing for hits.
Free · 25 credits a month · no card
How it works
- 1
Upload the clip. A 480p proxy is generated for analysis and for frame-accurate scrubbing, so you are not uploading a master to find out where the cuts are.
- 2
Choose a mode. General finds cuts and events across the whole clip; a focused mode such as footsteps goes after one class of hit and finds more of them.
- 3
Work from the result: audition an effect against any event, or hand the segment list to Scene Score and have music composed to it.
What you give it
One video file, a mode, and an optional tone hint. Analysis runs against the proxy, not the master.
What you get back
Scene segments, events with a timecode, a type, a label, a confidence and an intensity, dialogue regions, and suggested music cues. It is structured data, so the next tool can consume it rather than you retyping it.
WAV, FLAC and MP3, at 44.1 or 48 kHz and 16 or 24-bit. You pick the container and the rate on the form before you run it.
Who it is for
Editors sound-designing a cut on a deadline. Game cinematic teams passing a sequence for footsteps and impacts. Anyone who has scrubbed a two-minute clip frame by frame writing down timecodes on paper.
What it costs
Priced on the length of the clip, with the estimate shown before the run. Analysis is the cheap step; the generation it feeds is where the credits go.
Every run is a queued job. It holds its credit estimate while it works and settles to the real cost when it finishes; if it fails or you cancel it, the hold is released in full. You see the estimate before you spend anything.
Licence and ownership
Everything you generate is royalty-free for commercial use — client work, monetised video, games you sell, ads. No per-use fee and no attribution requirement. Generations are yours; uploads stay private by default, and the inputs to the isolator and splitter are deleted automatically.
Common questions
- What does the analysis actually return?
- Scene segments, a list of events each with a time, a type, a label, a confidence and an intensity, the regions where there is dialogue, and where music cues would fall. Confidence is included so you can ignore the guesses instead of auditioning all of them.
- What is the footsteps mode for?
- A general pass spreads its attention across everything in the clip and under-reports any one class. If what you need is every footstep in a two-minute walk, a focused pass finds them; a general pass finds some of them and a lot of other things.
- Can the results drive the other tools?
- That is the point of them being structured. The segment list goes to Scene Score to have music composed per scene, and any event can be handed to the sound effect generator as the moment a take has to land on.