How to find archive footage for documentaries (and stay copyright-safe)
Where documentary makers find licensed and public-domain archive footage, how to check what you can use, and how licence-aware search removes the biggest risk in long-form video.
Archive footage is what makes a documentary a documentary. It's also the slowest and riskiest part of production: hours of searching, and one wrong licence away from a copyright claim. Here's how experienced documentary makers do it.
Where the footage is
- Internet Archive (archive.org). The largest free collection: newsreels, government films, the Prelinger Archives of industrial and educational film. Licences vary by collection; check each item.
- Wikimedia Commons. Photos and some film, each with an explicit licence.
- NASA, ESA and NOAA. Space and earth-science imagery, almost all public domain.
- National archives. The US National Archives, the UK's Imperial War Museums (some restrictions), the Library of Congress, and equivalents in most countries.
- Prelinger and Periscope-type collections. Mid-century film with clear provenance.
- Universities and museums. Digitised holdings, often CC-licensed.
- Stock libraries. For modern establishing shots, not the subject itself. See stock footage for documentaries.
Reading the licence
| Licence | Safe for a monetised documentary? |
|---|---|
| Public domain / CC0 | Yes |
| CC-BY | Yes, with attribution |
| CC-BY-SA | Yes, with attribution; check share-alike obligations |
| CC-BY-NC | No: non-commercial |
| CC-BY-ND | No: you're editing it, which is a derivative |
| "All rights reserved" / unclear | No, unless you license it |
The trap on big archives: the licence field is filled in by whoever uploaded the item and is often wrong. Trust the collection's provenance more than the uploader's label.
A search routine that works
- Turn each narration line into two or three short search cores (place, event, year; object, action).
- Search collection by collection, not the whole archive at once.
- Log every asset: source URL, licence, collection, the line it's for.
- Download at the highest resolution available.
- Keep the log with the project. It's your defence if a claim arrives.
Budget 8–12 hours of this for a 10-minute film. It's most of the reason documentaries take so long.
Licence-aware search in Glisse
Glisse's footage scout does this routine automatically: it builds context-aware queries from each narration line, searches provenance-clear collections first, rejects NC and ND licences at search time, and keeps the source and licence for every asset. You pick from licence-badged results; the agent places them in the editor's template and cuts. The hours drop from ten to minutes, and the copyright risk drops with them.
FAQ
Where can I get free footage for a documentary?
The Internet Archive, Wikimedia Commons, NASA and national archives. Check the licence on each item, not just the collection.
Can I use YouTube videos as archive footage?
Not without permission. Most YouTube content is all-rights-reserved regardless of age.
How does Glisse find footage?
With a licence-aware search across provenance-clear archive collections, matched to each line of narration, then stock and generated footage for the gaps.
Make documentaries with an editor's taste
Glisse turns a real editor's style into a template. Describe your topic, and AI adapts it into a finished long-form film: licensed archive footage, maps, narration and music. Editors earn every time their template is used.
Join the waitlist →