Recording and audio

Perchnote records two things: your microphone and, optionally, your system audio (the sound coming out of your Mac, which is how it captures the other people on a call). Both get mixed into a single recording per meeting.

Permissions. macOS gates these behind privacy settings:

• Microphone. Required to record anything. You will be prompted the first time you record, or during onboarding.
• Screen Recording. Required to capture system audio. This is how macOS handles loopback audio capture, so even though Perchnote is not recording your screen, the permission lives under System Settings, then Privacy and Security, then Screen Recording. Without it, system audio comes through silent rather than throwing an error, so if the other side of a call is missing from your transcript, this is the first thing to check.

Audio settings (Settings, then Audio):

• Microphone device. Pick which input device to record from.
• Noise cancellation. Reduces background noise and echo on your mic input.
• Stereo recording. Puts you on the left channel and everyone else on the right, which helps speaker separation later.
• Accuracy pass. On by default. After a recording finishes, Perchnote does a full-quality second pass over the audio for a cleaner final transcript than the live one.

Recordings are saved as WAV files, one per meeting, named by the meeting's ID.

Never miss a recording. Perchnote watches which app is using your microphone. The moment Zoom, Teams, or your browser starts a call, you get a prompt to record it straight into the calendar event you are in. The watcher only sees the app name, never the audio.

Drop in an audio file. Drag any audio file onto the window (a Voice Memo, a phone-call recording, a podcast you want notes for) and Perchnote turns it into a fully transcribed meeting, complete with speakers and an AI summary if you have a provider configured.