Sort & tag
AI that watches every frame and sorts your whole shoot
Drop a folder of footage and Basecut watches and listens to every clip, ranks each one and tags it by what happens in the shot. A card dump becomes a ranked, searchable library before you sit down to edit.
Every clip ranked: hero, keeper, maybe, skip
Basecut gives every clip a verdict. Hero moments rise to the top, throwaways sink to the bottom, and dead footage goes to junk. Each verdict carries a star rating from 5 (hero) down to 1 (junk), so you start the edit from the keepers instead of scrubbing the whole card.
Tagged by what happens and how it was shot
Each clip is tagged by the kind of shot it is (interview, b-roll, action, establishing) and by how it was captured (drone, gimbal, tripod, handheld, vehicle). Those tags build the bins in your editor and feed the search in Basecut's library.
No naming, no spreadsheets
Point Basecut at the folder, start the run and come back to a sorted, tagged library you can send straight to your editor. The analysis runs on your own computer with Argus, Basecut's own AI. Basecut Cloud is an optional, faster engine on paid plans.
How a run works, step by step
1. Point Basecut at the card dump
Choose the folder that holds your cards or camera folders. Basecut reads the files where they sit and never changes them. Its results go into a run folder of their own.
2. Add a one-line brief if you want
The import screen has an optional brief box, up to 400 characters, used for that shoot only. Say what matters, such as the vows, the toasts or the first dance, and clips that match it rank higher.
3. Try ten clips, then run the rest
Before a large import, Basecut shows a time estimate and offers to try the first 10 clips, so you can judge the picks before you commit a night to the full run.
4. Basecut watches and listens
Each clip is analyzed for what happens on screen and what is said. Clips longer than 90 seconds are split into sections and judged section by section, so one long ceremony take can hold both a hero moment and a skip. If a run is interrupted it resumes without redoing finished clips, and you can retry any clip that failed.
5. Review the verdicts
Skim the junk and skip bins first to catch anything the AI underrated, then play through the heroes and keepers. If you disagree with a clip, set your own star rating. Yours wins over the AI's, and yours is the one that goes to your editor.
6. Choose the grouping and send
Set "Organise clips by" to shot type, rating or both (both is the default). Send everything or tick just the clips you want, and open the run in DaVinci Resolve, Premiere Pro or Final Cut Pro.
What you see on each clip card
Open a finished run in the Library and every clip card shows its rating and the reasons next to it, so you can decide quickly whether to trust it.
- Stars you can change
- The rating shows as one to five stars. Click the stars to set your own. When a clip has several sections, the card shows the best moment's rating and the time it happens.
- Warning tags
- Overexposed means blown highlights were found. Soft focus means the clip looks out of focus, and Soft shows the measured sharpness as a score out of 100. A tag such as Dead air, Restart take, Rambly or Fillers flags a problem with a spoken take. Check these clips before you use them.
- Brief match
- If you wrote a brief for the shoot, clips that match it carry a Brief match tag and rank higher.
- Check color
- When Basecut cannot tell whether a clip is log or normal video, the card says Check color, so you can confirm the log profile on the Color page.
- Filters and sorting
- A Filters panel narrows a run by verdict, by shot type or by a word you type. The sort menu offers Best first, Shot order, Longest first and File name.
How the verdicts show up in your editor
- DaVinci Resolve
- Resolve Studio, or free Resolve before 21.1, gets a project with bins, clip colors, star ratings and markers, plus a baseline log-to-Rec.709 grade on every clip. Free Resolve 21.1 and later gets a timeline file and .cube files instead.
- Premiere Pro
- A project with label colors, markers and a sequence for each verdict. Star ratings become label colors. No color is applied.
- Final Cut Pro
- FCPXML 1.11 with keyword collections. Four and five stars become Favorite, one and two stars become Reject. No color is applied.
Limits worth knowing
- The AI judges what it sees and hears. It does not know your couple or which relative the family will ask about, so treat the verdicts as a sorted pile with a recommendation, not as finished selects.
- On-device analysis needs an Apple Silicon Mac with 16 GB of memory or more, or a Windows PC with an NVIDIA graphics card with 10 GB or more of video memory. Other computers can use Basecut Cloud on a paid plan, or your own Gemini key.
- Native ARRIRAW and REDCODE R3D files are not supported yet.
Questions people ask
Does Basecut change or delete my footage?
No. It reads your original files and never modifies them. Results are saved in a separate run folder, and you can keep rejected clips as long as you like.
Does my footage get uploaded?
With Argus, no footage leaves your computer. If you choose Basecut Cloud on a paid plan, or use your own Gemini key, short reduced-size sections of your clips are sent to Google for analysis. Your original files are never uploaded.
Can I overrule a verdict?
Yes. Set your own star rating on any clip. Your rating wins over the AI's and is the one sent to your editor.
How long does a run take?
It depends on your computer and your footage. In Basecut's own tests on a 16 GB M4 Mac, a shoot of short clips took about two hours of processing per hour of footage, and a few long takes about one hour per hour. Plan on overnight for a full wedding. Basecut Cloud is the faster option.