HunterEye v2 · the voice-first way · follow the steps in order
The whole job in one line
Voice comes in → you cut it into parts by listening → for each part you find the matching footage → the tool names the files for you → you upload one pack to the Drive folder you are given.
You install NOTHING. Everything happens in the browser page.
1Open + sign in
2Upload the voice
Upload the narration audio file you are sent — not a script, the actual voice.
Keep the speech model on Tiny and press Transcribe the voice. Wait a little — you get the full narration written out, sentence by sentence, each with its time.
3Make one part at a time
Play the sentences and read them — understand what the voice is showing.
Tap the first sentence of the part, then the last sentence — or keep tapping to add more sentences. There is no limit: a part can be 1 sentence or 10. (Roughly 20–30 seconds of voice is a good size.)
Finish that part fully, then move to the next. Never all parts at once.
One partlisten + read it
→
Finish itfind its footage
→
Next partin order
4Search words
Under the part you see ready search phrases made from those exact sentences — names first (like “Voyager 1”).
Tap a phrase, or type your own words if you think of better ones. The suggestions are only wording help. YOU pick the footage — nothing is auto-picked.
5Hunt the footage
The tool searches 9 libraries at the same time and shows the clips as posters.
Library switches: the chips let you search only one library (only NASA, only Archive…) or all of them. “All” is the normal setting.
Same results again? Press Search again. Each extra press digs one page deeper in every library (up to page 3). The line under the part tells you how many clips you have and whether there is more to dig.
Nothing matches? Change the words and search again. There is no “empty scene” — the libraries hold almost everything; keep trying different words until it matches.
If a small grey note says some libraries didn’t answer, just press Search again — or switch that library off with its chip.
6Pick the winner — and the exact portion
Open a clip, watch it, and pick the one that truly matches what the voice is saying at that moment.
Need only a few seconds from a long clip? When you pick it, set the start and end (in/out). The full clip downloads once, and the wanted times are written into its file name — like P03_0m40-1m10. We cut the exact seconds on our side. You never cut anything.
When you tick a winner, the tool names the file for you by part number (P01, P02…). You never rename files yourself.
7Finish + upload
The progress line shows parts done / parts left.
At the end, download the pack (ZIP) — one file holding every picked clip inside it, each with its own name, plus the notes.
The finish screen tells you honestly which clips are INSIDE the ZIP and which are link-only — if any are link-only, download those files from their links too.
Upload the pack to the Drive folder link you are given. Done.
Your eye decidesThe tool suggests words. It never chooses footage.
One part at a timeIn order. Finish it, then the next.
Install nothingEverything happens in the browser page.
Stuck?Screenshot it and send it — don’t fight the tool.