Live translation3 min read
How to write role context for live translation
Separate audience, category, tone, and exact terminology into a maintainable translation role, then regression-test it against one real recording.
All articles
Every article, newest first.
22 articles
OBS & streaming tech3 min read
How to lay out bilingual live captions without hiding the product
Design bilingual live captions around language hierarchy, line length, contrast, placement, and real mobile previews instead of the OBS canvas alone.
Cross-border operations3 min read
A five-minute runbook for live translation failure
Classify viewer impact, isolate the failed layer, choose a safe fallback, and verify recovery from a viewer device without derailing the main broadcast.
Cross-border operations3 min read
A preflight checklist for translated live streams
Rehearse audio, recognition, translation, playback, captions, OBS, and the real viewer device with observable evidence before every broadcast.
Tool reviews3 min read
Local vs cloud speech recognition: choose by data flow
Compare local and cloud recognition by audio movement, network dependency, device load, and audit questions instead of treating privacy as a label.
OBS & streaming tech3 min read
Use OBS multitrack recording to audit live translation
Record host, translated speech, and programme mix on separate OBS tracks while retaining a complete track for ordinary playback and platform checks.
OBS & streaming tech3 min read
Captions but no translated audio in OBS: trace the signal
Trace silent translated playback across synthesis, virtual cable, OBS monitoring, stream tracks, and the platform instead of reinstalling blindly.
OBS & streaming tech3 min read
How to duck music under translated speech in OBS
Use OBS sidechain compression to lower music during translated playback, with verified starting values, test order, and misrouting checks.
Live translation3 min read
When translated speech falls behind: interrupt or queue?
Choose between interrupting, preserving, or clearing a live TTS queue based on content risk, with a repeatable stress test for translated playback.
Live translation3 min read
How to tune VAD silence timeout for live translation
Tune VAD silence timeout from observable failures such as split model numbers, merged sentences, and late finals, using one repeatable recording.
Tool reviews5 min read
How to test a live translation tool before you commit
Run every candidate against one fixed recording, compute word error rate, and read the error mix to locate the real problem. Plus latency and cost basis.
Live translation5 min read
Building a live translation glossary that actually works
Recognition-layer and translation-layer term lists have different limits. What the official docs say about phrase counts, boost values, and why one layer fails.