Video EditingIntermediate

How to Record and Edit a Multi-Track Podcast in Descript

Set up separate speaker tracks, label them, and edit each voice independently for a clean multi-person podcast.

10 minIntermediate

When two or more people talk, having each voice on its own track makes editing far easier: you can clean one mic without touching another, and crosstalk gets manageable. Descript supports multi-track recording and per-speaker editing. This guide covers a two-host setup.

What you need

Separate audio files for each speaker, or a multi-track recording session
A Descript project (audio-only or video podcast)
Speaker names ready so labels are clear

Step 1: Import each speaker as its own track

Add each person's audio file to the project. If they were recorded on separate mics or via a remote recording tool, you get isolated tracks. Descript stacks them so they stay in sync but remain individually editable.

Descript - Track view

Tracks

------------------------------------

Track 1 Host [||||--||||--|||---] Studio Sound: on

Track 2 Guest [--||||--|||--||||-] Studio Sound: on

Transcript labels speakers: Host / Guest

Cut on Track 1 does not affect Track 2

Two isolated speaker tracks aligned on the timeline.

Step 2: Label the speakers

In the transcript, assign a name to each track so the document reads Host and Guest instead of Speaker 1 and Speaker 2. Correct any spots where Descript attributed a line to the wrong person.

Step 3: Clean each track separately

Apply Studio Sound and noise reduction per track. The guest who joined from a laptop mic might need more cleanup than the host on a USB mic. Per-track effects let you treat each voice on its own merits.

Mute, do not delete, during crosstalk

When both people talk at once, instead of cutting, mute one track during the overlap so the dominant speaker comes through clearly. This keeps the conversation natural while fixing the muddy crosstalk.

Step 4: Edit the conversation as text

Now edit the combined transcript normally. Delete tangents, remove filler words, and tighten pacing. Because tracks are separate, removing the host's sentence will not chop into the guest's audio underneath it.

Step 5: Export the mix

Export an audio file (MP3 or WAV) for podcast hosts, or an MP4 if it is a video podcast. Descript mixes the tracks down into a single file with all your per-track effects baked in.

Descript - Labeled transcript

You

Host: So when did you first get into audio engineering?

Agent

Guest: Honestly by accident. I borrowed a friend's interface and never gave it back.

Speakers named so the transcript reads as a real conversation.

podcast-export

Format:  WAV (for the host platform) + MP3 (for distribution)
Mix:     Host + Guest tracks, normalized to -16 LUFS
Effects: Studio Sound per track, baked in
Chapters: optional, from transcript headings

Result: a two-host episode where the remote guest's rough mic was cleaned independently, crosstalk was tamed by muting, and the whole conversation was tightened by editing text, exported as a polished MP3.