Soundshine app icon

Soundshine

← Back to blog

How to Record a Podcast on a Mac

Recording a podcast on a Mac is mostly about one thing: getting your voice and your guest’s voice into a file at levels you can still fix afterwards. Everything else, including editing and publishing, has good free options.

The part macOS makes difficult is that your remote guest is not arriving through your microphone. They are arriving as system audio from Zoom or Discord or Riverside, and macOS does not let its own recording apps capture that.

What You Actually Need

A microphone that is not the built-in one. This matters more than any software choice on this page. The Mac’s internal mic sits above the keyboard and records the whole room, including your typing and the reflections off your desk. Any USB microphone is a dramatic improvement. Even wired earbuds with an inline mic beat it, because the capsule is near your mouth.

Headphones. Not optional. If your guest plays through speakers, your microphone records them acoustically at the same time your software records them digitally. Every word they say arrives twice, slightly out of step, and no amount of editing repairs it.

A quiet-ish room. Soft furnishings help. A bare room with hard walls sounds like a bare room with hard walls, and no plugin fixes that convincingly.

Software that records multiple sources separately. This is where the Mac’s built-in apps run out.

Why QuickTime and Voice Memos Do Not Work

Both record exactly one microphone input.

Solo, that is fine, and QuickTime Player will produce a perfectly good recording of your voice. With a guest it fails: you get your side of the conversation and silence where they should be, because they are system audio and system audio is invisible to those apps.

Recording Yourself and a Remote Guest

You want both sources in one recording, at levels you control independently.

  1. Open Soundshine Record and approve the recording permission once.
  2. Select your call app by name in the source list, such as zoom.us or Discord. Choosing the app rather than all system audio keeps notifications out of the episode.
  3. Add your microphone as a second source.
  4. Check both meters are moving before you start. Have your guest say something.
  5. Set the format to WAV or AIFF, because this is raw material headed for an editor.
  6. Record.

The reason to do it this way rather than with a single combined capture: every source has its own fader and live meter, adjustable while recording. Guests routinely arrive 10 or 12 dB quieter than the host. Instead of discovering that in the edit, you push their slider up in the first thirty seconds and the rest of the episode is balanced.

Passthrough monitoring lets you hear the mix as it records without it reaching your microphone, which is the echo problem solved in software if you forget your headphones.

Files write to ~/Music/SoundshineRecord. Requires macOS 15 or later, $14.99 one time.

Free and Professional Alternatives

GarageBand, free and already installed, is a real multitrack recorder and a decent editor. It records microphones and instrument inputs happily. Getting system audio into it still requires a virtual driver such as BlackHole, so it solves the editing half rather than the capture half.

Audacity, free, is the standard editor for spoken word. Same limitation: it cannot capture system audio on macOS by itself. Excellent as the second half of the workflow.

OBS, free, on macOS 13 or later, will capture your mic and a call app to separate tracks in one pass. Capable, and a lot of interface for an audio podcast.

Audio Hijack, $69, is the professional answer. Effects chains, per-source processing, scheduled recordings, silence splitting. If podcasting is your job rather than your hobby, it is worth the money. See Audio Hijack alternatives for how it compares.

Riverside, Zencastr and similar record each participant locally at full quality and upload afterwards, which sidesteps call compression entirely. Genuinely the best quality for remote interviews, and a monthly subscription.

Settings That Save You Later

48 kHz, stereo, uncompressed. Match what macOS runs at, and compress only at the final export.

Peaks around -6 dB. Leaves headroom. Anything touching 0 dB is clipped and unrecoverable.

Record a spare thirty seconds of silence. Noise reduction tools need a sample of the room to work from, and capturing it at the start costs nothing.

Clap once at the top. If anything ends up needing manual alignment, a sharp transient makes it trivial.

Test with a real two-minute take. Not ten seconds. Listen back on headphones before your guest’s calendar slot.

Record it in one file instead

Soundshine Record captures all your system audio, or just the apps you pick, plus your microphone, with per-source volume and no virtual driver to install.