Synced by construction
The caption lane reads transcript segments — no drift, no manual timing, ever.
Feature
Captions that look
designed, not generated.
Most feeds watch on mute — captions are the product, not decoration. Aperture generates them from the same transcript your edit uses, so they are synced by construction: cut the video and the captions cut with it.
The easy explanation
Your video already knows what it says.
Every upload is transcribed — every word, with timecodes and speaker labels. The caption lane reads that transcript. You never type timings. You fix a word in one place; it's fixed in the captions, the search, and the export — forever.
"…so the definately definitely worth it — you'll see the difference in week one."
click the word in the caption lane → fixed everywhere, forever
Click-to-edit
The list scrolls with the playhead. Click a line, fix the word, move on.
Style once
Font, size, color, position — set once per sequence, applied everywhere.
Burn in at export
The render bakes your styling in. Every platform sees your brand, not a default.
Titles & lower thirds
A text track, like any other lane.
Text is a clip type: type words, choose font, size, color, position, drag it on the timeline. Names, locations, calls to action — drawn live in the browser and rendered identically by the server. Safe zones keep everything out of the UI chrome on vertical exports.
Lower thirds
Names and roles, styled once, reused across the whole sequence.
End cards
Freeze-frame the last frame, drop a text clip, hold for the CTA.
Safe zones
Captions and titles compose inside vertical safe zones automatically.
Feature · Auto-captions
Watched on mute. Loved out loud.
Captions come from the same transcript as your edit — synced by construction, styled once, burned into every export your audience never hears.
Style your captionsWhat you get, plain.
- Transcription for every upload — speaker-labelled, searchable
- Auto-captions synced by construction — zero timing work
- Click-to-edit any word — fixes captions, search, and export
- Styled burn-in: font, size, color, position, per sequence
- Karaoke captions — each word highlights as it's spoken; word-level timings survive the export
- Caption style gallery at CC generation: clean / pop / boxed / karaoke
- Brand accent auto-applied to caption text from the tenant brand kit
- Text track for titles and lower thirds with live preview
- Vertical safe-zone guides on 9:16 and 1:1 exports — nothing hides behind platform UI
"Captions stopped being a chore when they stopped being a separate file."
— auto-captions, by constructionCaptions that look
designed, not generated.
From the same transcript as your edit. Styled once. Burned into every export.