Guide
A raw take is the recording exactly as it left the camera
A raw take is the recording exactly as it left the camera. Nothing removed, nothing trimmed: the pauses, the false starts and the sentence you restarted twice are all still in it. It is the input every edit starts from, and its quality sets the ceiling for the finished video. This page covers what a usable raw take contains and the habits that produce one.
- Continuous recording, restarts included
- 1
- The source ceiling per take in Cutroom
- 3 min
- Things no edit can add: audio quality and energy
- 2
Everything is still in it, and that is the point
A mistake inside a take is cheap. The edit removes it in seconds, and nobody watching the finished video knows it existed.
Material that was never recorded is the opposite. No edit can add a sentence you did not say or energy you did not have, so the raw take is the ceiling on everything downstream.
This is why experienced people keep rolling. Stopping the recording to fix a flubbed line costs a re-setup and the momentum. Restarting the sentence costs four seconds of file.
One take, start to finish
The only rule is that the camera does not stop. Everything else is repairable later.
Press record
Once
Talk
Flub a line, say it again
Leave a beat
Silence marks the joins
Stop
One file, everything in it
What a usable raw take contains
Close audio. Distance from mouth to microphone costs roughly six decibels per doubling, so a phone at arm's length beats a phone across the room by more than any setting can recover.
Restarted sentences, not restarted recordings. When a line goes wrong, pause for a breath and deliver the whole sentence again. The clean version survives and the broken one is cut.
A beat of silence between thoughts. Those gaps are where cuts land cleanly, and a take with no pauses in it forces every cut into the middle of a breath.
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
How long a raw take should run
Longer than the finished video, and not by a little. Three recorded minutes becoming a thirty second final is an ordinary ratio for spoken short-form, not a sign of failure.
There is also a hard number here. Cutroom accepts a source up to three minutes, transcribes every word of it, and cuts the pauses and false starts out in the first pass.
If the material genuinely needs more than three minutes of speech, it is not one short video. It is several, and each one deserves its own take.
Questions people ask
- Is a raw take the same as a rough cut?
- No. The raw take is the untouched recording. A rough cut is an early version of the edit made from it, with the structure in place and the polish still missing.
- Should I delete bad takes?
- Not before the edit. A take that felt bad often contains the best single sentence of the session, and storage is cheaper than re-recording. Delete after the final export ships, if at all.
- Do I need to clap or slate before speaking?
- Not for transcript-based editing. The transcript itself is the index, so any sentence can be found by its words rather than by a mark on a timeline.
Keep rolling, restart the sentence, and hand the mess over whole. Removing the mess is the mechanical part, and here it happens in the first pass.