The economics of owning a gold mine
A sleek, footage-forward business explainer on the real economics of a small gold mine — b-roll and data, end to end.
InfographicExplainerDocumentary b-roll
Documentaries with the real footage in them — found, placed, and cut for you.
When the script names a real person, place or machine, the video has to show it. Finding that clip, dropping it in, and trimming it to land on the right line is most of the edit. Framesail does that part: it finds the footage and photos, sets them in your channel's look, and cuts the whole thing to your voiceover — one finished video.
How the footage is found is on the AI b-roll generator page.
Framesailone video
The real footage, the real people, and the story around them — cut to one voiceover.
- No clip hunting
- No stitching or trimming
- One look across every shot
- A finished video, ready to upload
Write the script. Framesail places the real footage, draws what was never filmed, and cuts it all to your narration.
Made in Framesail
Documentary and explainer, real media inside.
Two finished videos, both with real footage and photos inside.
The most terrifying ways the world could end
A comedic whiteboard countdown through the end of the world — hand-drawn host, real photos, and a rising scale of dread.
WhiteboardHumorous countdownReal inserts
Where the time goes
The footage isn't the hard part. Getting it into the video is.
Every real thing your script names means a search, a clip or photo dropped into the timeline, a trim so it lands on the exact line that mentions it, and a pass to stop it looking pasted in. Do that for every beat and the footage is the whole edit.
The hunt
No more digging through libraries and archives.
Framesail finds real footage and real photos for every person, place and machine your script names, with the licence recorded.
The cut
Nothing to trim, nudge, or line up by hand.
Every clip and photo is placed and cut to the exact line of narration that mentions it, across the whole video.
The mismatch
It never looks pasted together.
Stock clips, archive photos and your own visuals usually arrive in three different looks. Here the real thing sits inside your channel's look, so the finished video reads as one piece.
You set the rule
You decide when the real thing appears. It never looks like filler.
Real footage shows up when your narration names a real person, place, or thing, and only then. Name three real things, get three. Agenda frames, transitions, and everyday objects stay drawn in your look, which is why the video never reads as a stock slideshow.
Two looks that switch it on.
Off by default. Pick one when you set up your channel, or tell your channel style in a sentence when you want the real thing on screen.
Documentary b-roll
Real footage carries the video: full-screen clips on every concrete moment, clean graphics in between.
Explainer with real inserts
Illustrated frames with the real thing set inside them: the actual logo, cover, or photo, in your look.
Real where it's real,
drawn where it isn't.
What one video got
54 seconds on Apollo 11, with the real Apollo 11 in it.
The real crew, the real Lunar Module, real launch-pad footage, and the stage-separation diagram nobody ever filmed, all in one whiteboard look, without opening a stock library once.
The first two frames are the same moment, before and after the real photograph is placed. Mission Control photo: NASA / Wikimedia Commons (public domain).
The honest version
The doubts a documentary maker has.
“Stock footage always looks like stock footage.”
It does when it's chosen to fill time. Here the real thing appears only when your narration names it, and never as decoration.
“AI can't show me the actual Neil Armstrong.”
It doesn't try. The frame is drawn with room left for the real photograph, and the real Armstrong is placed into it. Look at the Apollo frames above.
“What happens when it can't find the shot?”
The moment is drawn in your look instead, so nothing is left blank and the video still reads as one piece.
“I don't want to licence-check twenty photos.”
Footage and photos from Pexels are free for commercial use. Encyclopedic images carry their own licence, and Framesail records it with each one so you can check before you publish.
Questions
Questions about AI documentary video generator.
What is an AI documentary video generator?
A tool that makes a narrated documentary end to end — and can fetch the real footage and real photos a documentary depends on, not just draw an impression of them.
Where does the real footage come from?
Pexels for footage and photos, plus Wikipedia, Wikimedia Commons, and Open Library for the real people, places, logos, and covers your script names.
How do I turn real media on?
Pick the documentary b-roll look or the explainer-with-real-inserts look when you set up your channel, or tell your channel style in a sentence when you want real footage. It's off by default.
How much real media ends up in a video?
As much as the script names. Three real things, three inserts. No quota.
What happens when the b-roll can't be found?
That moment is drawn in your look instead, so the video is never left with a gap.
Can I use the fetched media commercially?
Pexels, yes. Encyclopedic images carry their own licence — check the source before publishing.
Does this work for explainer videos too?
Yes — the second preset is built for it. The illustrated look stays, the real logo or photo gets pasted in.
Does it replace generated visuals?
No. Diagrams, concepts and transitions were never filmed, so they are drawn in your look. Real people, places and machines are shown for real. Both land in the same finished video.
More on the main FAQ page. Both run on the same long-form AI video generator and script to video AI as every other video.
Show the real thing. Draw the rest.
The real footage, the real people, and the story around them — one finished video, in your look.