How to find a spoken phrase in VEGAS Pro

With Speech to Text

Speech to Text runs offline and supports more than a hundred languages, but it comes only with a VEGAS Pro subscription.

1. Select an event with speech and choose Tools › Speech to Text: the window opens in Transcript view.
2. Choose what to transcribe: Source Media (the whole source file) or Timeline Event (just the event on the timeline).
3. Pick the language in the Language list and a model in the AI Model list: Tiny to Large (roughly 78 MB to 3.1 GB), with Small as the default.
4. Click Analyze selected event or Analyze all.
5. Type the words into the Find box on the right of the window and step through matches with Previous and Next Occurrence.

Text-based editing

In the same window, the View list has a Text-based editing mode. Clicking a word moves the playback cursor there, and a range selected with Shift can be removed with Delete, cutting it from the timeline.

Where it gets awkward

Transcription runs on the events or sources you select, so a big archive has to be brought into the project and transcribed first. The Large model is more accurate but weighs over 3 GB and takes longer.

The same in Rescanzo

Rescanzo searches for the line across the whole footage folder at once and understands Russian and English speech. What it finds goes to VEGAS Pro as an EDL file.

1. Add your footage folder to Rescanzo and let indexing finish.
2. Switch to Pro mode and type the line, in Russian or English.
3. Add the results you want to a collection and choose the VEGAS Pro export: Rescanzo saves an EDL (CMX3600) with references to the source files and timecodes.
4. On Windows, Rescanzo launches VEGAS Pro and shows the file in File Explorer; import it from the VEGAS menu.

Questions and answers