How to Edit Zoom Recordings

Audio first

Fix the audio before you touch a single cut

A Zoom recording has one structural problem no amount of editing skill fixes after the fact: by default, everyone's audio gets mixed into a single track in real time, compressed for a live call, not for post-production. If cross-talk, a ringing phone, or a dog barking in the background survives the mix-down, it is baked into every speaker's audio for good. The fix happens before you record, not in the edit: in Zoom's desktop app, under Settings > Recordings, turning on "Record a separate audio file for each participant" gives every speaker their own track, so you can clean, level, and denoise each one independently instead of fighting a single flattened mix. The setting only applies to local recordings on the desktop app, up to 80 separate tracks, and it does nothing for anyone who calls in by phone, since phone participants share one track regardless.

Framing second

Decide what stays visible before gallery view decides for you

Zoom's gallery view resizes tiles based on who is talking, which looks fine live and terrible in a finished video, where a subject's frame should not jump size every time someone else unmutes. Lock speaker view during recording if the piece is a single-focus interview like a talking-head video, or plan to crop each participant's tile to a fixed frame in post if it is a multi-person panel, closer to editing a multicam interview. Either decision is easier to make before the call than after 90 minutes of footage exists.

Cut it like an interview

Treat the recording as a transcript, not a video file

Once the audio and framing problems are handled, a Zoom recording is functionally the same editing problem as any other interview: too much material, and a story buried somewhere inside it. Scrubbing through an hour of remote conversation to find the good parts is slow, and worse on Zoom footage specifically, because remote conversations run longer and looser than in-person ones. This is text-based editing: working from a transcript instead, reading the conversation and marking what is worth keeping, is the faster path regardless of where the footage came from. In ScriptCut, that transcript carries word-level timecodes, so a selected sentence maps to an exact frame in the recording, whether it is a single Zoom file or the separate per-participant tracks Zoom can output.

For a quick visual walkthrough of trimming, cropping, and captioning a Zoom recording before it goes anywhere near an NLE, this covers the basics:

Cross-talk

What actually fixes overlapping speakers

Cross-talk is the most common complaint about remote interviews, and separate audio tracks are the actual fix, not a workaround. With one mixed track, two people talking over each other is one unusable moment. With separate tracks, you can duck one speaker under the other, choose which voice leads, or cut around the overlap entirely, none of which is possible once two voices are baked into a single file. If you did not enable separate tracks before recording, cross-talk moments in the finished mix usually have to be cut rather than salvaged.

Alternatives

When it is worth recording somewhere other than Zoom

Zoom is fine for a casual conversation, but purpose-built remote recording tools solve the audio problem structurally instead of through a settings toggle. Riverside.fm and similar tools record each participant locally on their own device in uncompressed audio and up to 4K video, then upload progressively in the background, so a bad internet connection during the call does not degrade the final file the way it can on Zoom's real-time stream.

If remote interviews or podcast episodes are a recurring part of the workflow, switching the recording tool solves the audio problem once, and it is worth doing before you are batch editing a full season of them. If it is a one-off, cleaning up a Zoom file the way described above is usually enough.

Workflow

Start to finish, in order

Enable separate audio tracks in Zoom's recording settings before the call. Lock or plan framing for how the finished video should look, not how gallery view behaves live. Record, then pull the transcript (Zoom generates one automatically for cloud recordings, or run the local file through a transcription tool). Read the transcript, mark the material worth keeping, and cut filler words and dead air inline. Arrange the selected moments into order, clean each speaker's audio track individually where cross-talk or noise survived, and export the sequence to an NLE for final mix and any graphics.

Common mistakes

The three that show up most in Zoom-sourced edits

Recording without separate audio tracks and discovering the cross-talk problem only in the edit is the most common and the most preventable. Editing straight off the single mixed-down recording instead of pulling a transcript first wastes hours re-listening for quotes. And leaving Zoom's gallery-view resizing untouched in the final cut, so a speaker's frame visibly jumps size mid-sentence, is a small thing that reads as unpolished the moment a viewer notices it.

Sources

• Zoom Community: recording separate audio files per participant
• Riverside: Multitrack podcast recording, why it matters

Related reading: What Is Text-Based Editing?, How to Cut a Documentary Interview, and What Is Timecode?.