A short podcast video editing workflow with separate audio

The interesting moment in a recorded conversation rarely begins where the camera file begins. Someone answers a question, pauses, then reaches the point. To make a short video excerpt, you need to find that range, keep the good microphone steady and choose pictures that support the exchange.

The Studio preview is built for that bounded edit. Its current composition is limited to ten minutes, so use it for a short segment or a selected highlight, not an hour-long episode. If you only need written words from a full episode, the podcast transcript guide covers that separate task.

Choose the excerpt before arranging cameras

Listen through the source and note where the conversation becomes clear without a long introduction. Leave enough lead-in for the first sentence to make sense. At the other end, allow the thought to finish instead of cutting at the last keyword.

Bring the relevant video and separate microphone files into Studio. Local files can be dropped over the editor or chosen together. The editor initially places the first ten minutes of a longer source on the timeline, and you can set Source in and Source out in the Inspector to work with the range you need. Every valid source gets its own track. If a file fails to load, check its result; the other sources remain available.

One camera may have useful guide audio even if you plan to use a separate recorder as the final sound. Keep the originals nearby. Studio stores a browser draft, and a project backup contains edits and transcripts rather than the media bytes.

Establish the shared event

For an audio recording and two to four cameras, open Synchronise sources. Mark the same event in each original source, such as a clap or a clear spoken word. Studio can also suggest matching anchors from bounded local audio analysis. Review the proposed positions and play the sources together before applying the alignment as an undoable edit.

Audio matching during ordinary file import is a different, lighter action. It compares overlapping sound locally and moves only reliable distinct matches. Silence, short recordings, ambiguous matches or a position before timeline zero need your judgement. Neither workflow corrects clocks that drift gradually during a long recording. The browser sync walkthrough explains how to check a match at both ends of an excerpt.

Let the audio carry the cut

Choose the separate recording as master audio and listen for a natural start and finish. In the camera workflow, camera inputs exclude their own audio by default. Select shots on the picture-only programme track while the master recording continues underneath. A change of angle can cover a visual jump without interrupting a sentence, but listen again after every trim. The words should still connect.

Grouped sources move together until you explicitly detach one. Use a source-preserving trim where possible, then review the first and last frames of each chosen shot. Studio does not detect speakers or choose camera angles automatically. Those are editorial calls: a reaction shot may be more useful than a mechanically alternating view.

Add captions, then review suggestions

Select a source range to transcribe. Studio sends that audio to its transcription service and speech provider; transcription usage and the supported speech-to-text API-key setting apply. Timed words let you seek, and caption corrections can fix a name without changing the raw transcript. Check the words against playback, especially where people overlap. Export SRT or VTT mapped to your current timeline when captions are ready.

Literal filler search happens locally. Optional AI cut proposals use timed transcript words and require you to preview and accept each cut or a batch. Suggested highlights likewise need review; an excerpt can be previewed and exported without changing the saved composition. For a closer look at those controls, see transcript-based editing.

Export the whole short composition or a selected range as 720p video WebM. The range retains its mix and rebases sidecar captions to its own start. WebM export runs in real time with the tab visible. Save a project backup beside the original camera and audio files so the next edit can begin from the version you approved.