Transcription is fast, but proper nouns trip up every speech-to-text engine. Names of people, companies, locations, and technical terms are the most common source of transcript errors. Selects gives you a dedicated tool to catch and fix these before they end up in your export.
Where to find it
Open Check names and terms from the top bar in the workspace.
How it works
When you open the panel, Selects sends your transcript through an AI analysis that extracts proper nouns: people's names, brand names, locations, and technical terms. These are the words most likely to be misspelled by speech-to-text.
The panel displays the extracted list with each detected term shown alongside a Change To field.
How to use it
Review the list of detected terms. If a term is misspelled, type the correct spelling in the Change To field and click Update. The correction applies across your entire transcript, timeline, and export.
If a term is already correct, you can skip it. Focus on the ones that look wrong.
When to use it
Run Check names and terms after AI analysis completes and before you start editing. Fixing transcript errors early means your text-based editing, search, and exported transcript are all accurate from the start.
Why it matters
If you export a transcript with misspelled names, those errors carry into your NLE project, subtitle files, and any documents generated from the transcript. Fixing them in Selects is faster than hunting through a timeline in Premiere or Final Cut later.