Record a podcast or interview with remote guests

The problem

A standard video call is optimised for conversation, not for recording. It compresses audio heavily, drops quality when the connection wobbles, and gives you one mixed track where everyone is on top of each other. That is fine for a meeting and unusable as source material, because you cannot fix one person’s volume without changing everyone’s.

The specific failure that hurts most is the dropout. Somebody’s connection stutters for four seconds in the middle of the best answer of the interview, and the recording preserves the stutter permanently.

Then there is the work afterwards. An hour of raw conversation contains maybe twenty usable minutes, and finding them means listening to the whole thing again.

What AI can actually do here

The recording fix is not AI at all and it is the most important part: recording each participant locally at full quality and uploading the files afterwards, so the network affects the conversation but not the recording. If the connection stutters, the saved track is still clean.

Separate tracks per participant is the second half of that. It means you can raise one quiet guest without raising the room noise on everyone else, which is the difference between amateur and acceptable.

The AI layer handles the editing. Transcription lets you edit by text, removing a tangent by deleting a paragraph. Filler word removal takes out the ums without you finding each one. Automatic clip generation proposes the moments worth cutting for social.

Voice tools cover the repairs: fixing a word that was mangled, or producing a version in another language. Use these carefully on interviews, because altering what a guest appears to have said is a different act from tidying your own audio.

How the tools differ

Riverside is the recording studio. Free gives 2 hours of multi-track recording as a one-off, at 720p with a watermark and 44.1 kHz. Pro is $29 per month or $24 billed annually, with 15 hours of separate track downloads per month, up to 4K, no watermark, text-based editing and Magic Clips. Grow is $39 or $34 annually with 20 hours plus social scheduling, hosting and live streaming. The Webinar tier at $99 or $79 annually adds up to 100 registrants.

Descript is where the editing gets serious once the recording exists. Free covers 60 minutes of transcription per month; Hobbyist $16 per user per month annually, Creator $24, Business $50. If you are already editing there, its recording features may cover you without a second subscription.

ElevenLabs is only relevant for voice work: dubbing into another language or generating narration. Free is 10,000 credits per month without a commercial licence, Starter $6 per month with one.

The practical split: use the first for capture, the second for assembly, the third only if you need voice generation.

When you don't need a tool

If you publish raw conversations to a small audience who care about the content rather than the polish, a normal call recording is fine and free. Podcast production values are a cost, not a requirement.

If you record in person in one room, a single decent microphone beats any remote recording setup.

And if the output is short clips rather than a full episode, see turn long videos into short clips and skip most of this.