How to Capture and Transcribe System Audio on Mac
System audio capture — transcribing whatever audio is playing on your Mac — is one of the most requested features for speech-to-text tools. Traditionally, this required installing virtual audio drivers like Soundflower or BlackHole, which can be complicated to set up, cause audio conflicts, and require maintenance across macOS updates. Modern solutions capture system audio natively. Here is how to set it up.
Step-by-Step Guide
Understand the traditional approach (and its problems)
Historically, capturing system audio on Mac required a virtual audio driver (Soundflower, BlackHole, Loopback) that creates a fake audio output device. You route Mac audio to this virtual device, then record from it. This is fragile — drivers break with macOS updates, can interfere with other audio apps, and require kernel extensions or system extensions.
Use a native system audio capture solution
Glasscribe uses Apple's ScreenCaptureKit API to capture system audio directly, without virtual audio drivers. This is the same API that macOS uses for screen recording. It is stable, updates with macOS, and does not interfere with your audio routing. Install Glasscribe and the system audio capture works out of the box.
Grant the necessary permissions
The first time you use system audio capture, macOS will prompt you for Screen Recording permission (since ScreenCaptureKit requires this). Go to System Settings > Privacy & Security > Screen Recording and enable Glasscribe. You may need to restart the app after granting permission.
Select System Audio as input source
Open Glasscribe from the menu bar and select "System Audio" as your input source. You will see this option alongside your microphone inputs. Choose the appropriate language for the audio you want to transcribe.
Start capturing and transcribing
Begin the transcription session. Any audio playing on your Mac — from any application — will be captured and transcribed in real time. This includes browser audio, media players, video calls, streaming services, and any other sound source.
Pro Tips
Frequently Asked Questions
Does system audio capture require a virtual audio driver?
No. Glasscribe uses Apple's ScreenCaptureKit API, which provides native system audio capture without any third-party drivers. This means no kernel extensions, no audio routing configuration, and no compatibility issues with macOS updates.
Can I still hear the audio while it is being transcribed?
Yes. System audio capture works alongside your normal audio output. You hear everything through your speakers or headphones as usual, and the transcription runs simultaneously in the background.
Why does system audio capture require Screen Recording permission?
Apple's ScreenCaptureKit API — which provides system audio access — is governed by the Screen Recording permission. This is a macOS security measure to ensure apps cannot secretly capture your screen or audio without your knowledge. Granting this permission allows Glasscribe to capture system audio only.