Talon/voice-cat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(audio): stereo mic capture on Windows & macOS desktop clients
Some checks failed
Build Linux Binaries / linux/amd64 (push) Has been cancelled
Details
Build Linux Binaries / linux/arm64 (push) Has been cancelled
Details

Browse Source 
Both desktop mics were hard-mono: the core defaults capture_channels=1 and
neither client ever called vc_set_capture_channels (only iOS did). Add a
persisted "Stereo microphone" toggle to each client's Audio settings, applied
when the mic stream starts and live via vc_set_capture_channels + vc_audio_restart.
Expose both ABI calls in the Windows interop; the macOS wrapper already had them.

Core fix: encode_and_send_frame now folds a stereo mic frame to mono on a mono
channel - previously the channels==2 branch encoded interleaved L/R directly even
on a mono channel, feeding a mono opus_encode 2x its samples (wrong pitch/garbage).
Real stereo still only reaches the wire on a stereo channel; on a mono channel the
mic is cleanly downmixed.

Test: test_stereo_mic_mono_channel. ctest --preset dev green (28/28). Docs: voice.md.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Talon
2026-06-23 20:48:26 +02:00
parent b14cf2a4e8
commit f72219ddf3
11 changed files with 210 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
clients/apple/macOS/VoiceCatMac/Windows/MainWindowController.swift
View File

@@ -62,6 +62,7 @@ final class MainWindowController: NSWindowController, NSWindowDelegate {
static let vadThreshold = "voice.vadThreshold"
static let inputGain = "voice.inputGain"
static let inputNoiseReduction = "voice.inputNoiseReduction"
static let stereoMic = "voice.stereoMic"
static let pttKeyCode = "voice.pttKeyCode"
static let auxEnabled = "voice.auxEnabled"
static let auxDeviceUID = "voice.auxDeviceUID"
@@ -80,6 +81,12 @@ final class MainWindowController: NSWindowController, NSWindowDelegate {
internal var inputNoiseReduction: Bool = false {
didSet { UserDefaults.standard.set(inputNoiseReduction, forKey: AudioDefaults.inputNoiseReduction) }
}
// Capture the mic in stereo (interleaved L/R) instead of mono. Real stereo only reaches the
// wire on a stereo channel; the core folds a stereo mic to mono on a mono channel. Applied to
// the core when the mic stream starts (micToggleClicked) and live via SettingsWindowController.
internal var stereoMic: Bool = false {
didSet { UserDefaults.standard.set(stereoMic, forKey: AudioDefaults.stereoMic) }
}
internal var selectedInputDeviceId: String?
// Aux outgoing stream: a second hardware input device the client captures itself and feeds to
@@ -163,6 +170,7 @@ final class MainWindowController: NSWindowController, NSWindowDelegate {
if d.object(forKey: AudioDefaults.inputNoiseReduction) != nil {
inputNoiseReduction = d.bool(forKey: AudioDefaults.inputNoiseReduction)
}
stereoMic = d.bool(forKey: AudioDefaults.stereoMic)
if d.object(forKey: AudioDefaults.pttKeyCode) != nil {
pttKeyCode = UInt16(d.integer(forKey: AudioDefaults.pttKeyCode))
}
@@ -781,6 +789,9 @@ final class MainWindowController: NSWindowController, NSWindowDelegate {
if let devId = selectedInputDeviceId {
client.setInputDevice(streamId: streamId, deviceId: devId)
}
// Stored on the stream before the announce round-trip completes, so the core's
// first capture-device open (ensure_audio_running) picks up the channel count.
client.setCaptureChannels(streamId: streamId, channels: stereoMic ? 2 : 1)
client.setInputMode(selectedInputMode)
if selectedInputMode == .voiceActivation {
client.setVadThreshold(vadThresholdValue)