Best footage logging software for post-production (2026)

StoryFolder — 2026-09-02

Last updated: 2 September 2026. Every price and capability below carries the date it was last recorded. The figures marked 2 September 2026 were read from the vendor's own page that day; the rest were recorded on 1 September 2026 and have not been re-checked since. Confirm current price and capability with the vendor before you buy anything.

You can search a transcript, find the sentence where someone says "and that's when the whole thing changed," and still not be able to find the silent establishing shot of the building, the client-specific b-roll clip, or which of six drives it's actually sitting on. Footage logging is not one problem, and no tool here solves all of it. We publish this comparison and StoryFolder is one of the entries, so weigh the recommendations accordingly. Every price and capability below carries the date it was last recorded, and we concede plainly where StoryFolder doesn't fit.

No single tool logs footage well for every job. StoryFolder is the shot-level visual index for one owner on a local machine; Fast Video Cataloger locates clips on offline Windows drives; Kyno inspects media for an NLE handoff; iconik and axle.ai run shared, governed storage; Airtable is a manual log for a disciplined team.

Say the limit early: StoryFolder is not a media asset manager. It doesn't track which drive your footage lives on, doesn't index a folder you haven't imported, doesn't sync across a team's shared storage, and has no NLE round-trip. If any of those is the actual job, a catalogue or a MAM in the table below wins it outright, and one real account shows what StoryFolder is for instead. Filling in metadata by hand is normal work: in our own analysis of 295 paying accounts, recorded 1 September 2026, one account had the same three fields (Client, Job# and Title) filled in 3,852 times each, 11,556 entries in total, across dozens of boards. That's someone typing the same three things over and over because there was no better place to put them, which is exactly the gap a local shot-level index is for.

What kind of footage logging problem do you have?

Before picking a tool, work out which of six jobs you actually have, because "footage logging" covers all of them and no product covers all six:

1. You need a schema: consistent fields like client, job number, location, typed per clip or per shot.
2. You need to find something by what's in the picture: a visual index, searchable by description or tag.
3. You need to find something by what was said: transcript or speech search.
4. You need to know which drive it's on: an offline-volume locator.
5. You need a team to share one library: shared storage with permissions and governance.
6. You need the result inside an NLE: media inspection that hands off to an edit.

This page is about logging footage after it's shot. On-set camera and continuity logging is its own category with its own tools. If your job is (1) or (2) on a machine you own, keep reading past the next section. If it's (4) or (5), skip to the software built for those jobs specifically; a shot-level index is the wrong tool for either.

Which software is best for a local footage archive?

For one person or a small team working off their own machine, the split is: StoryFolder for shot-level visual metadata and search, Fast Video Cataloger for knowing which offline Windows drive a clip lives on, and Kyno for inspecting media and moving selections toward an NLE.

StoryFolder's library search covers Text, Dropdown, Tag and Number fields, on a shot or on the video, plus the transcript, in one query, and it's a case-insensitive substring match where every word of a multi-word query has to match. Rating and Checkbox fields are reached through filter chips rather than the search box. It does not do fuzzy matching, and it does not search filenames or file paths. What it does do is useful: type a client name and get back both the shots someone tagged with it and the moment someone said it on camera. It's also not what "AI visual search" implies to some people, so don't go in expecting semantic search across footage nobody described. If a file gets moved to a new location, relink reconnects the source without losing any metadata already on the shots. Since 0.4 the same box reaches the transcript because transcription runs locally on import, and the library has collections for hand-picked shots and quick filters for the queries you rerun. Local transcription · the library controls.

Fast Video Cataloger works the other way round: its own order page promises you can "Search and browse even when video files are disconnected", across catalogues of 10,000-plus videos, so the drive can be sitting unplugged on a shelf when you search for what's on it. What that search actually matches on isn't stated there. Kyno, per its vendor pages, browses local and NAS media in place, reads and writes metadata, cuts subclips and hands selections and their metadata toward an NLE, which is the job StoryFolder does not do at all. We have not run any of the three against shared footage, so the table's "unverified" labels stand.

Which software is best for a shared team library?

If simultaneous users, shared proxies, role-based permissions or real media governance are requirements, look at iconik or axle.ai, not StoryFolder. iconik's Starter plan runs from a free Collaborator tier through Browse, Standard and Power seats, priced per active user per month, with pay-per-use AI credits on top, and its Professional and Enterprise plans are quoted by sales (iconik.io/pricing, read 2 September 2026). That's the cost structure of a real MAM. axle.ai publishes a cloud offer at $20 per terabyte per month, an on-premise MAM from $2,995 and its Tags product from $200 a month (axle.ai, read 2 September 2026), a storage-first way to think about the same job. StoryFolder's flat monthly or annual price reflects the opposite choice: it doesn't store or manage your original media centrally, so there's no usage dimension to price.

Airtable sits in an odd spot here: genuinely collaborative, with views and automations, but it's a database sitting beside your footage rather than a video-aware tool. There's no shot detection, no media-aware playback, and no timecode; you're building the log structure yourself and pasting in links.

Do transcripts solve footage logging by themselves?

No, not for anything silent. The open-source arkiv project says it indexed a "1,506-clip real production library (1,161 of them dialogue-free B-roll)", which is 77%. Those figures were read off the project's repository on 2 September 2026; the README doesn't say when that library was indexed, and the most recent test results it posts are dated 22 May 2026. That's one real library rather than a claimed industry average, but a useful gut check on how much of most footage a transcript alone will never reach. Transcription finds what's said. It does nothing for the establishing shot of the parking lot, the b-roll of hands on a keyboard, or anything else nobody talks over.

StoryFolder's transcription runs on-device via Whisper, free, and your footage is never uploaded to be transcribed. AI Vision Autofill is a separate feature: a paid, cloud-based call that sends a representative frame from each shot to a third-party vision model to fill in metadata fields you've turned on. Keep those two apart, and the help page on what goes where spells out the split. Neither one is a semantic search over the footage itself. What you're searching, in both cases, is text that transcription or a person or Autofill actually wrote down.

Can AI log footage using your own categories?

Yes, with real limits worth knowing before you rely on it. StoryFolder's AI Vision Autofill can fill in the metadata fields you've enabled, using your own field names rather than a generic tag vocabulary, but every field starts with Autofill switched off, so nothing gets touched until you turn it on per field. Dropdown, Checkbox, Rating and Number fields are constrained to their own valid values; it cannot invent a new dropdown option. Tag and Text fields are open-ended by design, since a text field's whole point is that it can hold anything a person could type, which also means that's exactly where a model can write something plausible and wrong, like a job number nobody typed.

A value outside a field's valid options is silently dropped rather than flagged, and there's no undo for an Autofill run specifically. There is no published accuracy figure for it, so judge it on your own footage before you rely on it. Two global switches shape how far it reaches, and the full Autofill contract sets out the rest: Only fill empty fields, which never overwrites something a person already typed, and Autofill on import, which is off by default, because running a vision model automatically over every shot in a long import would be hundreds of calls nobody asked for.

Two other tools in the table sell AI on this axis, and they sell a different shape of it. iconik's pricing page lists transcription, object and scene detection, facial recognition and translation, bought as AI credits on top of the per-seat price. axle.ai's own site lists trainable faces, scene understanding, logo and text recognition, spoken-word search and vector search across its cloud, on-prem and Tags products. Neither page says whether those results can land in field names you defined yourself, which is the axis this section is about. Unverified: check the vendors' own documentation before comparing either vendor's AI tagging with ours. Both capability lists above were read from those vendors' own pages on 2 September 2026. Airtable, Fast Video Cataloger and Kyno are not marketed on automatic visual description on the pages cited here.

What should a useful footage log export?

Whatever tool you use, the export needs to carry: a stable identifier per clip or shot, a thumbnail or preview, start/end/duration, the descriptions or categories someone entered, transcript context where it exists, source or location information, and a route into whatever comes next, whether that's an edit, a client deliverable or an archive record.

StoryFolder's spreadsheet export gives you a thumbnailed XLSX and CSV together, plus a PDF, per-shot images and per-shot clips, but the timecode in that spreadsheet is elapsed from the file's start rather than source timecode, and there's no NLE round-trip on any of it. Airtable's export is only as good as what your team typed in, since it has no media-aware fields at all. Kyno and other NLE-oriented tools are stronger exactly where StoryFolder is weakest: getting metadata and selections to actually land inside an edit. A MAM's export is stronger again where shared originals and proxies are involved, which is a different job from the one a one-owner shot index is built for.

Which footage logging tool should you choose?

1. Tiny archive, disciplined operator, no video-native needs: Airtable or a spreadsheet.
2. One owner, shot-level visual index on your own machine: StoryFolder.
3. Windows, need to know which offline drive holds a clip: Fast Video Cataloger.
4. Local or NAS media inspection with an NLE handoff: Kyno.
5. A team needs shared, governed storage: iconik, or axle.ai's published per-TB tier.

Whatever you pick, run it on one real project first: time how long it takes to bring footage in, describe it, find one silent b-roll shot again by search, reconnect a source that's moved, and hand a select to an edit. That tells you more than any table does.

How do these tools compare at a glance?

We have not run a shared benchmark file across these tools, so treat any claim about which logs footage fastest or most accurately as unverified until someone does.

FAQ

What's the cheapest Windows offline-video catalogue here? Fast Video Cataloger, at $9.90/month, $97/year, or $197 perpetual, one Windows PC per licence.

Can StoryFolder search dialogue and shot tags in the same query? Yes. One library search covers Text, Dropdown, Tag and Number fields plus the transcript together, using substring matching rather than fuzzy matching.

Does StoryFolder search disconnected hard drives? No. Relink reconnects a source that's moved, but locating a clip on an offline volume is Fast Video Cataloger's category, not StoryFolder's.

Is StoryFolder's AI logging on-device? No, not the part that logs metadata. Transcription is on-device and free; AI Vision Autofill is a separate, paid, cloud call that sends representative frames to a third-party vision model.

How many field types can StoryFolder log with? Six: Text, Tag, Dropdown, Checkbox, Rating and Number. Every field's Autofill switch starts off.

Is iconik a lightweight desktop app? No. It's a shared MAM whose Starter plan runs $9 to $120 per active user per month plus pay-per-use AI credits, with a $0 Collaborator tier for view-only access.

Does a transcript alone log all of a video's footage? No. The arkiv project's repository, read on 2 September 2026, reports a 1,506-clip production library with 1,161 clips of dialogue-free b-roll, 77%, and none of that gets found by transcript search.

← All articles