How to record a podcast with a remote guest without betting the episode on one connection

Apple Podcasts publishes one loudness target for every episode in its catalog: around -16 dB LKFS, with a tolerance of a single decibel either way, and a true peak that never goes above -1 dB FS. No video call hands you audio that lands there on its own. So the answer to how to record a podcast with a remote guest is not a piece of software, it is a method: each person's microphone is recorded locally, on their own machine, as its own file, and the call is treated as the channel you talk over rather than the thing capturing the audio. Production people call it a double-ender. It works with one guest and it works with four, and it is the only approach that survives a bad connection, because the recording never travels across the connection in the first place.

Any show can be listed on PodAnswer for free, and a listed show gets its own page that can turn up when someone searches the show by name, so if you are recording guest episodes and want them findable outside the podcast apps, you can add your show at https://podanswer.com/signup.

What happens to a voice between the guest's mouth and your hard drive

Conferencing software is built for intelligibility across a connection that changes second by second, and that goal produces a signal chain working against you. Echo cancellation runs before the audio leaves the guest's machine, and it ducks whichever side is quieter, which is why laughter and interruptions come back thin. Noise suppression removes room tone, and it takes the first syllable of a sentence with it when the gate closes early. Automatic gain control rides the level up and down, so a guest who leans back gets amplified along with their refrigerator. Then the codec throws away whatever the bandwidth cannot carry, and packet loss leaves gaps no editor repairs, because there is nothing underneath the gap to recover.

Zoom's documentation includes an original sound option that switches off echo cancellation and automatic gain control, and turning it on genuinely improves a call recording. It does not convert a call recording into a local one. The compression and the packet loss are still in the path.

A local file skips the entire chain. The microphone writes to the guest's own disk at full rate, the connection quality stops mattering to the recording, and the file you edit is the file the microphone made.

Running a double-ender so the guest does not have to produce anything

One. Pick where the guest's file gets written before the session, not during it. A browser studio handles this invisibly: the guest clicks a link, grants microphone access, and their track records to their machine while the conversation happens. The alternative is asking the guest to open a recorder on their own computer, which works for a producer or another podcaster and fails for a cabinet secretary's press office.

Two. Record your own side locally too, in your own interface or recorder, started before you open the call. Hosts forget this one and end up with a pristine guest track against their own compressed call audio.

Three. Record the call as well. Zoom, Riverside, Meet, whatever you are using, leave its own recording running. That file is worthless for broadcast and invaluable on the day a local track does not arrive.

Four. Clap once at the top, both of you. The transient gives your editor a single sharp point to line the tracks against, which beats dragging waveforms by eye.

Five. Confirm the files exist before anyone leaves. Keep the guest on the line while the upload finishes or while they drag their file into the folder you sent. A guest who has closed the laptop is a guest you are now emailing.

Six. Ask the guest to plug in headphones and nothing else. Headphones remove the bleed of your voice into their microphone, which does more for the finished audio than any other single change and is the only piece of equipment worth insisting on. The rest of their setup you can fix in post.

If you are still building a guest list to run this process on, the four places that actually produce bookings are covered in how to find podcast guests, and the reverse direction, getting yourself booked on other shows, is in how to get booked as a podcast guest.

The step that eats episodes is the upload, not the recording

Local recording and a file in your possession are two different events, and the gap between them is where remote episodes die. Zencastr's published plan comparison lists local recording, progressive uploading and automatic cloud backups as three separate features, which is an honest description of the problem. A browser studio writes the guest's track to their disk and then pushes it to the service in chunks while the conversation continues. If the guest closes the tab at the end of a ninety-minute interview, the chunks that have not gone yet may never go. Progressive upload exists so that most of the file is already safe by the time the call ends, and it is worth checking that whatever tool you pick does it, rather than uploading in one block at the finish.

Three habits cover the rest. Check the recording indicator at the ten minute mark and say out loud that you are doing it, which also gives the guest permission to flag a problem. Keep the call recording as the fallback, since a compressed track that exists beats a lossless track that does not. Download the separate tracks the same day, because entry tiers cap how long files are held.

The specifications on the other end of the pipeline

The reason to protect the local file is what the distribution end expects. Apple publishes its requirements rather than leaving them to guesswork, and they are the targets worth editing toward.

For audio in an RSS feed, Apple Podcasts accepts MP3 or AAC, and recommends 64 to 128 kbps for a mono file at 44.1 or 48 kHz, and 128 to 256 kbps for stereo at the same sample rates. Apple states plainly that it strongly recommends AAC over MP3 for feeds, and that within AAC the MP4 container beats ADTS for streaming efficiency and accurate seeking. For subscriber audio uploaded directly through Apple Podcasts Connect, the accepted formats are WAV, FLAC or MP3 with a floor of 44.1 kHz at 16 or 24 bit, and single-channel WAV and FLAC files are refused outright.

The loudness figures are the ones most shows miss. Apple asks for about -16 dB LKFS within one decibel and a true peak no higher than -1 dB FS, both calculated by the ITU-R BS.1770-5 method, with the conditioning done before encoding rather than after. A guest track captured through a conference codec with automatic gain riding it cannot be brought to that standard reliably. A local track at 48 kHz can. The two figures, what the meters are counting when they report them and the order to apply them in are set out in how loud should a podcast be.

In practice: record and edit at 48 kHz, export mono AAC at 96 to 128 kbps for a two-person interview unless music is doing real work in the show, and measure loudness rather than eyeballing the waveform. File size and format also feed the analytics, since what counts as a podcast download is decided partly by how much of the file a player pulls.

What the recording studios charge, and what the free tiers stop at

The browser studios have converged on a similar shape, and their free tiers suit a test rather than a season. Riverside's published plan page puts its free tier at two hours of multi-track recording, 720p video, 44.1 kHz audio and a watermark on the output. The 48 kHz audio and the separate-track downloads sit on the paid tiers, with Pro listed at $288 billed annually and a 14-day trial, and the number of separate-track download hours rising with each tier above that. Zencastr's comparison table similarly splits 44 kHz MP3 recording from 48 kHz WAV recording across plans, so the sample rate your show records at is a billing question at both companies.

That is the real decision. A two-person audio show that wants a 48 kHz separate track per person pays somewhere in the low hundreds a year, and the alternative is free: both people record locally in whatever they already have, a voice memo app included, and you handle the files by hand. The free route costs you an email thread per episode.

For the craft underneath all of this, the engineers who do it for a living are worth hearing directly. Recording Studio Rockstars works through microphone technique and mix decisions episode after episode, including the source-versus-processing question that decides a remote interview, as in how do you get a deep punchy snare sound, where the answer turns out to be about capture rather than plugins.

When the guest cannot manage any of it

Some guests will not install anything, will not click a studio link, and will join from a phone in a car. The fallback that holds up: ask them to record a voice memo on the phone itself, propped close, and send it afterward. A phone microphone at close range beats a laptop microphone at conversational distance by a wide margin, and the file is clean enough to edit. Align it to your call recording using the clap and the episode sounds like an interview rather than a speakerphone.

Record your own side properly in that scenario, since you are now carrying the audio quality of the whole episode on one track. For how long the finished thing should run, the hosting data is in how long should a podcast episode be, and if you are also moving the show between hosting companies while you change your recording workflow, the things that break during a migration are listed in can you switch podcast hosts.

Short answers

What is the best way to record a podcast with a remote guest? Separate local tracks for every participant, a call recording kept as backup, a clap for alignment, and headphones on the guest. The software choice matters far less than those four things.

How do I record a podcast with remote guests, plural? The same way, with one file per person. Browser studios price this as separate-track or multi-track recording, and the number of participants allowed is a plan limit at every service, so check the cap before you book a four-person panel.

How do I record a podcast remotely for free? Have each person record locally with software they already own, including a phone voice memo, and collect the files afterward. The free tier of a browser studio works for a test or two, but the caps on recording hours, sample rate and separate-track downloads make it unsuitable for a weekly show.

Can I just use Zoom? You can, and the result is a compressed single-stream recording with echo cancellation and automatic gain baked into it permanently. Turn on the original sound option if that is your route, and keep in mind that nothing in post recovers what the codec discarded during the call.

Will listeners notice a remote interview? A double-ender is indistinguishable from an in-room recording to almost every listener, because the audio never crossed the internet in real time. A call recording is recognizable as one immediately.

What to take away

Record locally, one file per person, every time. Keep the call recording as insurance. Clap at the top. Insist on headphones. Confirm the files exist before the guest hangs up. Edit toward -16 dB LKFS and a -1 dB FS true peak, and export mono AAC.

PodAnswer turns podcast episodes into answer articles that rank for the questions listeners type into Google, each excerpt linking back to the show. How that works is at https://podanswer.com/for-podcasters, and the Growth plan at four articles a month and the Network plan at twelve are at https://podanswer.com/pricing.