Troubleshooting

EchoPad cannot hear the system audio. Open Settings → Recording → Permissions and allow System Audio Recording. After rebuilding EchoPad from source macOS can treat it as a new app; if the switch is on but nothing is recorded, turn it off and on again in System Settings → Privacy & Security → Screen & System Audio Recording.

If it still fails, set Capture with to ScreenCaptureKit.

macOS asks for permission when recording starts, and audio before you answer is not delivered. Stop, then start again.

Your voice leaks from the speakers into the microphone, or from the microphone into the call. Headphones fix it. EchoPad removes most of this echo, but not all.

Speaker detection works best with at least a few seconds per person. Rename speakers in the conversation to merge them in the transcript. For a known number of people, Transcribe Again after the call often helps.

Setup needs an internet connection to fetch the models from Hugging Face. Check the connection and press Try Again in Settings → Transcription.

Another app may use the same combination. Pick another in Settings → General.

Allow notifications for EchoPad in System Settings → Notifications, and check Settings → General → Notify me when a call starts. Detection watches which app uses the microphone, so it starts when you join, not when the app opens.