Introducing Enhance: One Click to Broadcast-Quality Subtitles
Anyone who has ever worked with subtitles knows the grunt work. The transcription is done, every word is right — and now the real job begins: splitting the long blocks, recalculating timecodes, inserting gaps, breaking lines where the sentence actually pauses, hunting down the three different spellings of the same product name. Block by block, line by line.
Enhance does that job in one click. It takes your entire raw subtitle file — not one block at a time, all of it — and restructures it to professional, standardised form. You review the result, adjust anything you like, and export.
This article shows exactly what it does, with a real example, and explains the credit system that powers it — including why every plan, free included, comes with enough credits to try it on your own footage.
What Enhance actually does
Raw automatic transcription gives you accurate text segmented by pause detection — which is to say, segmented by where the speaker happened to breathe. Enhance re-segments it by the rules professional subtitlers work to. In one pass, across the whole file, it:
- Splits long blocks into short, readable units — each with its own recalculated, synchronised timecodes
- Breaks lines at natural pauses in the sentence, following semantic logic rather than a character count
- Inserts industry-standard micro-gaps between consecutive blocks, so the eye registers each new subtitle
- Reduces reading speed where a block carries too much text per second for a viewer to finish
- Applies one consistent spelling of names and brands across the entire file
- Optionally tidies spoken language — removing filler words, and repairing the grammar that spontaneous speech leaves dangling
That last group is worth a concrete illustration, because it mirrors how people actually talk:
Raw: "It's, um, it's basically a model that was developed to, you know, visualize things a little."
Enhanced: "It's basically a model that was developed to visualize things."
Same statement, same meaning, same speaker — minus the scaffolding that belongs in the room but not on the screen. And because this crosses the line from formatting into wording, it's a setting you control: run Enhance verbatim-strict, or let it tidy. (For when each mode is the right call, see the honest verbatim-versus-edited discussion in our subtitle quality guide.)
Before you run anything, you choose a preset — or set every parameter yourself. After it runs, the editor is exactly where you left it: every block open to manual adjustment. Enhance does the ninety percent that is mechanical; the judgement stays yours.
A real example, not a demo reel
This is an actual block from an English-language podcast processed in Inwista — nearly twenty seconds of speech that the raw transcription delivered as one block of roughly 300 characters:
BEFORE — 1 block, 19.7 seconds 00:46:51 → 00:47:11 Yes, and I'm realizing that our conversation has been a little recipe-focused as far as cooking is concerned. So when we're saying pasta, we're usually talking about wheat-based dough and sauces, and when we're talking about rice dishes, we're often talking about grains, stock and vegetables, and that's not always the case.
Every word is correct. As a subtitle, it's a wall — five-plus lines squatting on the picture for twenty seconds.
One click of Enhance later, it's six blocks, each with its own synchronised timecodes, natural gaps between them, and line breaks placed where the sentence pauses. The first three:
AFTER — 6 blocks (first three shown) 00:46:51 → 00:46:56 Yes, and I'm realizing that our conversation has been a little 00:46:56 → 00:46:59 recipe-focused as far as cooking is concerned. 00:46:59 → 00:47:04 So when we're saying pasta, we're usually talking about wheat-based dough
…and so on through the remaining clauses, down to a tidy closing block.
Doing this by hand means finding five split points, calculating five new pairs of timecodes, and placing ten line breaks — for one block. A forty-minute episode contains hundreds of them. That's the arithmetic that makes Enhance less a convenience than a category change: the difference between "we should really clean up the captions" and captions that are simply clean.
What Enhance is not
Two honest boundaries, so expectations land where the product does:
It's "nearly perfect" — a deliberate phrase. Enhance clears the mechanical work: the splitting, timing, gapping, breaking. It does not replace your read-through. You review the result in the editor, and occasionally you'll rephrase a block or nudge a split — that's the design, not a shortfall. The tool's job is to make your review the only remaining work.
It restructures; it doesn't rewrite. Outside the optional spoken-language tidying — which you switch on knowingly — Enhance changes the form of your subtitles, not the substance. What was said remains what was said.
Credits: how Enhance is paid for
Enhance runs on AI credits — and this is the part worth thirty seconds of your attention, because it's separate from your transcription allowance.
- Transcription hours cover turning audio into text: 4 uploads of up to 20 minutes on Free, 10 hours a month on Lite, 25 on Pro.
- Credits power AI features on top of that transcription — Enhance chief among them.
Every plan includes credits, refilled automatically each month — including the free plan. Free accounts get 100 credits, Lite 1,000, Pro 2,500, Enterprise by agreement. We set the free allocation deliberately high enough for anyone to run Enhance on their own real footage — because a before/after on your own video is more convincing than anything we could write here.
Two practical notes:
- Your remaining balance is visible at the top right of the editor at all times.
- If Enhance ever returns an error after several runs, the file is fine — the credits are empty. Nothing is broken; the balance refills automatically at the start of your next month, or immediately on upgrade.
How to run it
- Upload a file with + New Project — or open any project you've already transcribed.
- When transcription finishes, hit Edit. You land directly in the Enhance window.
- Pick one of the ready-made presets, or set the parameters yourself — including whether spoken-language tidying is on or off.
- Click Apply changes. A moment later, the whole file is restructured.
- Review and adjust in the editor, exactly as before.
- Export — as an SRT or VTT subtitle file, or as video with the subtitles burned in. (Not sure which you need? We've written the decision guide.)
Why this matters beyond aesthetics
Clean subtitle structure isn't cosmetic. Reading speed, block length, gaps and line breaks are the substance of professional standards — the ones streaming platforms enforce at delivery and the ones accessibility law quietly assumes. The European Accessibility Act's captioning requirement is satisfied by captions that are accurate, synchronised, complete and readable — and readable is precisely the property raw auto-captions lack. Enhance is, in effect, the readability requirement automated.
The full craft — what the standards are, and why each exists — is in What Makes a Good Subtitle?. Enhance is that article, applied in one click.
Try it on your own footage
Inwista's free plan lets you run the whole workflow on your own material — upload, transcribe, structure, edit and export. Current limits and plan details are on the pricing page.
Upload a video and run Enhance →
Frequently asked questions
Does Enhance change what I said? Its core work — splitting, timing, gaps, line breaks, consistent spelling — changes structure, not wording. The optional spoken-language setting can remove fillers and repair dangling spoken grammar; you decide whether it runs, and everything remains editable afterwards.
Which plans include Enhance? All of them. Every plan receives monthly AI credits — 100 on Free, 1,000 on Lite, 2,500 on Pro, by agreement on Enterprise — refilled automatically. Credits are separate from your transcription hours.
I got an error running Enhance. Is my file broken? Almost certainly not — an error after several runs typically means your credits for the month are spent. The balance shows top-right in the editor and refills automatically next month, or instantly on upgrade.
Can I undo or adjust what Enhance did? Yes. The result lands in the editor with every block open to manual change, exactly as before. Many users run Enhance first and then make a handful of judgement edits — that's the intended workflow.
Does it work on languages other than English? Yes — it was built handling Norwegian's spoken forms and dialectal spellings, which are considerably harder than English's, and the examples in this article are from an English-language podcast processed in Inwista.
How is this different from just transcribing? Transcription answers "what was said." Enhance answers "how should it sit on screen" — segmentation, timing, gaps, line breaks, reading speed. They're different problems; most tools solve only the first.