Editing captions and speakers

Every clip is built from the word-level transcript. Editing the transcript is how you fix on-screen captions, rename speakers, and trim a clip's start or end.

Editing caption text

Open a clip and click the caption you want to fix. The transcript editor highlights the word and lets you:

Correct a misheard word (proper nouns, brand names, acronyms).
Add or remove punctuation.
Split or join words.

Changes save automatically and apply to the burned-in captions on the next render. After editing, click Re-render clip to produce a new MP4.

Renaming speakers

If your source has two or more speakers, Deepgram labels them Speaker 0, Speaker 1, and so on. To rename:

Open any clip that contains the speaker.
Click the speaker label in the transcript.
Enter the speaker's name (e.g. "Alex" or "Guest").

The rename propagates to every clip in the project. This matters most when you're using the split-screen layout — the speaker name appears on the active panel.

Adjusting clip start and end

To trim the edges of a clip, drag the transcript handles at the start and end. The clip duration updates live. You can't extend past the highlights the model originally selected by more than a few seconds; if you need a longer segment, delete the clip and add a custom range from the source timeline.

Fixing framing

By default, clips center-crop the vertical frame. If face detection is enabled for your workspace, the crop follows the active speaker automatically.

You can override framing on a single clip by choosing a layout:

Single — one speaker fills the frame.
Split screen — two speakers stacked vertically. Only available when the highlight contains at least two speakers.

Re-rendering