Skip to content

fix: exclude FaceTime/avconferenced from audio tapping to prevent volume ducking#157

Open
IftatBhuiyan wants to merge 2 commits intoronitsingh10:mainfrom
IftatBhuiyan:fix/facetime-audio-ducking
Open

fix: exclude FaceTime/avconferenced from audio tapping to prevent volume ducking#157
IftatBhuiyan wants to merge 2 commits intoronitsingh10:mainfrom
IftatBhuiyan:fix/facetime-audio-ducking

Conversation

@IftatBhuiyan
Copy link
Contributor

@IftatBhuiyan IftatBhuiyan commented Mar 10, 2026

Problem

When FineTune is running, FaceTime call audio (incoming audio from the other person) is extremely quiet. This "ducking" effect makes calls unusable. Quitting FineTune restores normal levels.

Root Cause

AudioProcessMonitor was detecting avconferenced (the real-time audio daemon for FaceTime) and creating a process tap with muteBehavior = .mutedWhenTapped. This intercepts the audio stream and breaks macOS's hardware Acoustic Echo Cancellation (AEC) loop. The system detects the broken AEC loop and drastically ducks the volume as a fail-safe to prevent feedback.

Fix

Added FaceTime and avconferenced to the system daemon filter lists in AudioProcessMonitor.swift so they are ignored by the tapping engine:

  • systemDaemonPrefixes: com.apple.FaceTime, com.apple.avconference
  • systemDaemonNames: FaceTime, avconferenced

Note: This is a workaround that preserves audio quality at the cost of FaceTime not appearing in the FineTune app list.

Testing

  • FaceTime call audio remains at normal volume with FineTune running.
  • Dictation still works correctly (chime plays, activation reliable).
  • Other audio apps (Music, browsers, etc.) remain controllable.

Fixes #156

…h10#123)

macOS dictation chime was suppressed and dictation sometimes failed to
start because FineTune's AudioProcessMonitor was not filtering
dictation-related system daemons. These processes were being tapped with
muteBehavior=.mutedWhenTapped, intercepting their audio output.

Add the following to systemDaemonPrefixes:
- com.apple.SpeechRecognitionCore (speech-to-text daemon)
- com.apple.speech (speech framework services)
- com.apple.dictation (dictation UI/coordination)
- com.apple.corespeech / com.apple.CoreSpeech (CoreSpeech daemon)
- com.apple.VoiceControl / com.apple.voicecontrol

Add the following to systemDaemonNames:
- speechrecognitiond, dictationd, corespeech

Fixes ronitsingh10#123
…ume ducking

When FineTune creates a process tap on avconferenced (FaceTime's real-time
audio daemon) with muteBehavior=.mutedWhenTapped, it breaks the hardware
Acoustic Echo Cancellation (AEC) loop. The system detects the broken AEC
and drastically ducks audio volume as a fail-safe against mic feedback.

Add the following to systemDaemonPrefixes:
- com.apple.FaceTime
- com.apple.avconference

Add the following to systemDaemonNames:
- FaceTime
- avconferenced
@IftatBhuiyan IftatBhuiyan marked this pull request as draft March 10, 2026 03:22
@IftatBhuiyan IftatBhuiyan marked this pull request as ready for review March 10, 2026 21:21
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[Bug] FaceTime call audio is extremely quiet when FineTune is running

1 participant