Enhancing subtitles

Raw transcription gives you accurate text — enhancement turns it into professional subtitles. In one operation, Inwista's AI restructures your captions to follow the rules broadcasters live by: sensible line breaks, comfortable reading speed, clean timing. You choose the standard; the AI does the tedious part. You'll find it behind the wand icon in the editor's left sidebar.

Start with a preset

The Enhancement Preset selector sets every rule at once to match a publishing context:

• Standard (Default) Balanced settings for general use — two lines, 42 characters per line, 15 characters-per-second reading speed.
• Social Media (TikTok / Reels) Built for phones and fast cuts: single-line captions, 35 characters, faster reading speed, no gaps between blocks, and blocks capped at four seconds.
• TV & Streaming (Netflix Standard) The strict broadcast rulebook: two lines of 42, 15 CPS, hyphens separating speakers in dialogue.
• Custom Adjust any individual rule yourself — the preset switches to Custom automatically the moment you change one.

Layout & Appearance

These rules shape how each subtitle block looks. Max Lines Per Block (the broadcast standard is 2) and Max Characters Per Line (usually 42) set the frame. Block & Line Balancing controls the silhouette of two-line blocks — the classic choice is bottom-heavy, the "pyramid" shape that reads most naturally.

When one sentence must split across blocks, Continuation Markers can signal it — a dash or an ellipsis, at the end of the first block, the start of the second, or both. And when two people speak in one block, Speaker & Dialogue Format separates them: hyphens (the standard), the speaker's name, or the name in brackets.

Language Rules

Text Condensation is the one rule that may change your words. None (Verbatim) keeps every word exactly as spoken. Smart and Aggressive allow the AI to rewrite and tighten sentences when the dialogue is too fast to display within your reading-speed limit — trading word-for-word fidelity for readability. For interviews and legal content, stay verbatim; for fast-talking social clips, Smart earns its keep.

Timing & Pacing

Max Reading Speed caps how fast text flashes by, in characters per second — 15 CPS is the industry standard, and it's the rule most raw transcriptions break. Minimum and Maximum Block Duration keep blocks on screen between 1 and 7 seconds by default, so nothing blinks past or overstays through a silence.

Gap Between Blocks inserts a brief empty frame between consecutive subtitles so the eye registers the change — from none at all to the SDH standard (~231 ms), with broadcasting (~99 ms) and streaming (~132 ms) in between.

Cost, running it, and undoing it

The panel shows an estimated credit cost before you commit — it scales with how much subtitle text your project contains. Click Apply changes and the enhancement runs on the whole project; it takes a moment for longer videos, and the result loads straight into the editor.

Two safety notes. First: enhancement automatically saves your pre-enhancement state, and the result arrives as a revision labeled Enhanced — if you don't like the outcome, the Revisions panel takes you right back (see Revisions & rollback). Second: the rules can work against each other — a very low reading speed plus verbatim text plus short max duration leaves the AI no room to satisfy all three. If results look strained, loosen a rule rather than stacking stricter ones.