How to record your Mac screen with system audio

Record your screen on a Mac, play the file back, and the sound is missing — not your microphone, but the thing you actually wanted: the app’s audio. This is macOS being careful, not broken. Here’s why it happens and the three ways around it, from a free virtual‑device dance to a recorder that treats system audio as a first‑class track.

Why the sound is missing by default

macOS treats the microphone and the system’s output as two different things with two different rules. The built‑in Screenshot toolbar (Cmd‑Shift‑5) offers your mic under Options, but no system audio — the Mac never routes what your speakers are playing into a screen recording. It’s a privacy posture as much as a limitation — silent lifting would let every call and every notification leak into a recording.

QuickTime Player’s screen recording is the same story: microphone yes, system audio no. So the demo where the app plays a sound, the call where you want the other side’s audio, the game with its soundtrack — all of them come out silent unless you add something.

Method 1: a virtual audio device (free, and fiddly)

The classic fix is a virtual audio driver such as BlackHole or Loopback. The driver creates a fake output device; you set the Mac (or just the app) to play through it, and the recorder treats the fake device as an input. The sound reaches the recording instead of your speakers — usually alongside a second mic track for your voice.

It works, it’s free, and it has a failure mode worth naming: while the virtual device holds the audio, your headphones hear nothing unless you also set up a multi‑output device in Audio MIDI Setup. Every demo day starts with a trip to Audio MIDI Setup, and every once in a while a meeting starts silent because the setup reverted.

Method 2: QuickTime with the workaround

QuickTime can record system audio only through that same virtual device — there’s no built‑in toggle. If you’re already comfortable in Audio MIDI Setup, the recipe is: install the driver, point QuickTime’s microphone setting at it, and accept that you won’t hear the audio live. For a one‑off capture where the silence during recording doesn’t matter, it’s fine.

Method 3: a recorder with system audio built in

Capture apps that target demos treat system audio as a checkbox, because a demo without the app’s sound isn’t a demo. Snapbak records the Mac’s audio output as its own track — separate from your microphone and your camera — so you can keep both, drop one, or balance them later instead of guessing at mix levels while recording. The toggle is right in the recording controls: system audio on or off, per recording.

The practical difference shows up on playback. A demo recorded with one merged audio track is locked into whatever the levels were live. Separate tracks mean the app’s sounds can stay crisp while your narration dips under them, or get dropped entirely when the footage becomes a GIF and audio leaves the picture anyway.

Which should you use?

One‑off recording where you control the audio chain: the virtual device costs nothing and works. Regular demos, tutorials, or bug reports where sound matters: a recorder with system audio as its own track removes the Audio MIDI Setup ritual entirely — that’s the job Snapbak was built for, and the download is free for the first 100 users.

One rule applies to all three: recording a call captures other people’s voices, and the consent rules for that don’t change with the method — the short version lives in the meeting‑recording post.