Find the clip where someone said something
The job isn't "organize my videos." The job is "where is the clip where they said X?" You remember the sentence, not the filename or the timestamp. Here's how to find it — starting with the honest manual methods that are already on your Mac, then how Archivist Sort does it locally when the archive is too big to scrub by hand.
Start with the honest manual methods
Before any tool, and regardless of whether you ever buy anything, you can find a lot of footage with what's already on your Mac. These are the manual passes that work for a small, well-named archive:
1. Search the file names first
Finder search on your footage folder, by the terms you remember. If you were disciplined about naming — "client-interview-release-plan.mov" — the filename often contains the answer. This is triage, not proof, but it's free and it's usually first.
2. Read the notes and logs around the footage
Interviewers keep call sheets, shot logs, transcripts, and notes. A searchable text file next to a video file can point you at the right clip far faster than scrubbing it. Look for companion documents in the same folder before you reach for anything heavier.
3. Scrub the likely candidates
If you've narrowed it to a handful of clips, open them in QuickTime and scrub. For a short archive this is the honest, fastest method — no tool will beat your eyes on five files.
When manual stops scaling
The moment your archive outgrows scrubbing, the bottleneck isn't watching — it's that the footage has no searchable words. A folder structure can't answer "which clip is the release plan?" because that information lives in speech, and speech isn't in the filename.
At that point you want transcription: turning each clip's spoken words into searchable text. You could transcribe each file by hand with a terminal tool, but doing that across an entire archive, one file at a time, is the same exhaustion you were trying to avoid.
How Archivist Sort searches your video
Archivist Sort transcribes your archive locally with Whisper-family backends and makes every spoken word searchable across the whole archive — not file by file. The app finds Homebrew and pipx installs of whisper-cli, openai-whisper, or mlx-whisper automatically, so you don't manage a path or a server.
Everything stays on your Mac. No cloud, no accounts, no upload — the source files and their transcripts never leave your machine.
Transcription is optional
Sort works fully without transcription. If you only need the file-cleanup and archive-organizing core, your footage is never touched. Transcription is something you opt into because you want speech searchable.
The app is honest about its limits
Transcription quality follows your backend and your hardware, and the app never invents a transcript. Settings shows exactly which binary was found, where, and why — or the specific reason none is usable. If there's no working backend, it tells you, rather than silently producing an empty or fabricated result.
Searching and getting the clip out
Once your footage is transcribed, search on the Report screen for the phrase you remember. Results come from across the entire archive, and duplicate hits collapse to unique clips — the same moment that appears in two source files doesn't clutter your results.
When the search finds the footage you need, export the matching source files as an experimental FCP7 XML (xmeml) bin — a handoff starting point for Resolve or Premiere, not a finished timeline. Two honest limits worth knowing before you rely on it:
- Whole files, not trimmed ranges. The exporter hands you the matching source clips in full. Trimming each hit to its exact range is on the roadmap.
- It is not FCPXML. The exporter is honest about its dialect — importers expecting the newer format will not be silently handed the wrong one.
So the workflow is: transcribe locally, search every spoken word, pull out the whole matching source clips, and finish the assembly in your editor. It's a search-and-handoff tool, not a finished-edit generator — and it's honest that it isn't. Reading your archive is free — download, transcribe, and search at no cost; $29 once is executing a cleanup plan through Sort, with quarantine, receipts, and byte-verified rollback.
Frequently asked questions
Does my footage leave my machine?
No. Transcription runs on your Mac, and the app makes no calls to the internet for it. Source files and transcripts stay local.
Do I need a specific backend installed?
Yes — one of the Whisper-family backends. Archivist Sort finds Homebrew and pipx installs of whisper-cli, openai-whisper, or mlx-whisper automatically, and Settings reports exactly what it found and where.
Does video search change or delete my files?
No. Searching is read-only on your footage, and exporting a bin copies out the matching source files — nothing is altered or deleted.
What if I only wanted file cleanup?
Then you can ignore transcription entirely — Sort works fully without it. The cleanup and archive-organizing core runs the same whether or not you ever transcribe a frame.
The bigger picture
Video search is one feature of a local-first, review-first cleanup tool. It follows the same spine as everything else: inventory, review, approve, quarantine, rollback — and it never invents, never uploads, and never executes without your acknowledgement.
See the full product page and its video archive intelligence section: Archivist Sort video search.