Localize a webinar Short: add captions first, dub only if it earns it
Localizing a webinar Short for new markets: start with translated captions at $0.20 a language, and only dub where captions don't get the views.
The cheapest way to localize a webinar Short is to burn translated captions first: one Video captions job per language at $0.20 for a clip up to 60 seconds (Sume docs, read 2026-10-03). Dubbing replaces the speaker's voice, takes a transcription, translation and speech step, and brings approval questions about whose voice it is. Start with captions, watch the results per market, and dub only where captions alone do not work.
Decide per market
Captions keep the real speaker and add no synthetic voice; for a one-minute clip, a reader can follow along. Markets where viewers rarely read subtitles may need a dub. For the pipeline see the dubbing pipeline post; for cost see dubbing cost per minute.
| Question | Captions | Dub |
|---|---|---|
| Keeps the speaker's real voice | Yes | No |
| Needs a translation | Yes | Yes |
| Needs a voice choice and consent | No | Yes |
| Sume cost per Short | $0.20 per language | Multiple steps; see the cost post |
Process
Keep one clean master of the Short. Get each translation reviewed by a fluent reader. Send each as cues or script_text and set the style to match the script: omitted style follows the wording, not the language. Then run the QA checklist on every version. A translated copy of the same clip may still be treated as minimal change by platforms, so add a market-specific intro of your own.
Sources
Related posts
More in Use cases
- Luma event cover image: square, 800 px minimum, text off the corners
Luma recommends a square 1:1 event cover at 800 by 800 px or more, static over GIF. How to generate one with Sume that survives rounded corners and small cards.
- Lunar New Year 2027 greeting video: a plan from 126 days out
Lunar New Year falls on Saturday 6 February 2027. Build a 15-second greeting from six stills, a music bed and caption cues for about $0.62 in Sume costs.
- Lyria 3.5 section timestamps: write a 30-second ad bed prompt
Lyria 3.5 reads section timestamps like [0:00 - 0:10] in the prompt. On Sume, put them in the Music Router prompt and read the section map from result.lyrics.
- Mailchimp video merge tag: the thumbnail links out, so pick it well
Mailchimp video merge tags show a thumbnail that links to a URL, so the still is the ad. Choose the frame with Sume video frames: 1 to 24 stills, source size.
Written by Sume