Remote Recording Setups That Sound Professional

Most podcast conversations now happen across a video call. The good news is that remote episodes can sound as polished as studio ones. The bad news is that the default setup, recording the call itself, almost never does. Here is the setup professionals use, and the habits that keep it working.

Record a double-ender, not the call

A double-ender means each person records their own voice locally, on their own device, and the files are combined afterwards. The call is only for talking. The recording never touches the internet.

Why it matters:

• Call audio is compressed to survive bad connections. It sounds thin and hollow and it cannot be fixed later.
• A dropped connection on the call does not damage the recording.
• Each voice lands on its own track, so you can edit one person without touching the other.

The trade-off is a little coordination. Your guest needs to press record too, and you need to collect their file afterwards. It is worth it every time.

Choose a local recording app

You have 3 options, in rough order of effort:

• Browser-based studios such as Riverside, SquadCast and Zencastr. The guest clicks a link, the app records locally in the browser and uploads in the background. The least effort for guests. Ask them to use Chrome and to stay on the page until the upload finishes.
• A desktop recorder on each side. Audacity, GarageBand, QuickTime or a DAW. More reliable than a browser, but your guest has to be comfortable with the software.
• The phone voice memo app as a last resort. Surprisingly good through a decent headset, and almost everyone knows how to use it.

Whichever you choose, record a backup. The browser studio records locally, you record your own side in a desktop app, and the call is recorded as a safety copy. Three layers, zero panic.

Headphones, always

Every participant wears headphones. No exceptions, no "my speakers are quiet".

Without headphones, the other person's voice plays out of a speaker, into the mic, and back into the recording as echo. Software echo cancellation fights this by cutting bits of the audio, which is why open-speaker calls sound choppy. Headphones remove the problem entirely.

Wired earbuds are fine. Over-ear is better. Bluetooth is the weakest choice because it adds latency and often forces the mic into a low-quality mode.

Keep the sample rate consistent

This one causes more mysterious problems than any other setting. If one person records at 44.1 kHz and the other at 48 kHz, the tracks will slowly drift out of sync over a long episode. A 60-minute conversation can end several seconds apart.

• Pick one rate, usually 48 kHz, and ask everyone to use it.
• Record in WAV at 24-bit if there is space, otherwise 16-bit. Avoid MP3 for raw recordings.
• Check the setting in the app and in the system sound settings. They can disagree.

If you receive mismatched files, convert them to the same rate before you line them up.

Plan for the internet to fail

Assume the connection will drop at some point. Then design around it.

• Ask guests to use a wired connection or sit close to the router.
• Close video, if you can bear to. Audio-only calls survive weak connections far better.
• Agree a rule in advance. If the call drops, both people keep recording, count to 5 out loud and pick up from the last question. The counts become your sync points later.
• Have a phone number ready as a backup line.

The local recording keeps rolling through all of it. That is the whole point of the double-ender.

Sync the tracks

With two or more files in hand, lining them up is quick if you prepared for it.

• Start every session with a clap. Everyone claps once, on a count of 3, with their mic in front of them. The clap shows up as a sharp spike in each waveform and you align the spikes.
• Most editors can auto-align by waveform, but the clap is your safety net.
• Check the end of the episode as well as the start. If the end has drifted, you have a sample rate mismatch.

Trim the pre-roll, keep the clap as a reference until the final export, then cut it.

Level mismatched guests

The most common remote problem is not noise. It is two voices at different volumes, recorded on different microphones in different rooms. One is a broadcast dynamic mic in a treated office. The other is a laptop in a kitchen.

You can fix this by hand with compression and EQ on each track, and it is worth learning. But if you are publishing weekly, let the tools help. On Witspeak Podcasts, AI Voice Studio removes background noise, levels each speaker to the same loudness and masters the episode to -16 LUFS, which is the standard podcast apps expect. One tap, and the kitchen laptop sits comfortably next to the studio mic.

A simple pre-call checklist

Send this to guests the day before:

• Headphones on, wired if possible.
• Quiet room, soft furnishings, door closed.
• Phone on silent, notifications off on the computer.
• Chrome open, no other tabs doing heavy work.
• Water nearby.
• We will clap at the start. Please stay on the page after we finish so the upload completes.

Guests who get a checklist arrive calmer and sound better.

Ready to publish

Once the tracks are synced and levelled, upload to the studio at podcast.witspeak.com and your episode goes out to Apple Podcasts, Spotify, YouTube Music, Amazon Music and the rest from one RSS feed, free. Create your show and send your next guest the checklist today.