feat(audio): stereo mic capture on Windows & macOS desktop clients

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>

clients/apple/macOS/VoiceCatMac/Windows/MainWindowController.swift

@@ -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)

clients/apple/macOS/VoiceCatMac/Windows/SettingsWindowController.swift

@@ -63,6 +63,12 @@ final class SettingsWindowController: NSWindowController, NSWindowDelegate {
     private let nrCheckbox = NSButton(checkboxWithTitle: "Noise reduction (RNNoise)",
                                        target: nil, action: nil)
 
+    // 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. Persisted
+    // via MainWindowController.stereoMic.
+    private let stereoMicCheckbox = NSButton(checkboxWithTitle: "Stereo microphone",
+                                             target: nil, action: nil)
+
     // Aux input stream: a second outgoing stream from another hardware input device (e.g. line-in
     // / aux), captured client-side. Device + volume only — aux is always-on. Persisted via
     // MainWindowController.auxEnabled / auxDeviceUID / auxGain.
@@ -178,6 +184,11 @@ final class SettingsWindowController: NSWindowController, NSWindowDelegate {
         nrCheckbox.setAccessibilityLabel("Microphone noise reduction")
         nrCheckbox.setAccessibilityHelp("RNNoise denoising of your microphone. Cleans your signal for everyone.")
 
+        stereoMicCheckbox.target = self
+        stereoMicCheckbox.action = #selector(stereoMicChanged)
+        stereoMicCheckbox.setAccessibilityLabel("Stereo microphone")
+        stereoMicCheckbox.setAccessibilityHelp("Capture your microphone in stereo. Only transmitted in stereo on a stereo channel.")
+
         let inputModeRow = NSStackView(views: [inputModeLabel, inputModeControl])
         inputModeRow.orientation = .horizontal
         inputModeRow.spacing = 8
@@ -259,7 +270,7 @@ final class SettingsWindowController: NSWindowController, NSWindowDelegate {
         volumeRow.orientation = .horizontal
         volumeRow.spacing = 8
 
-        let stack = NSStackView(views: [inputModeRow, vadRow, inputGainRow, nrCheckbox, pttRow, deviceRow,
+        let stack = NSStackView(views: [inputModeRow, vadRow, inputGainRow, nrCheckbox, stereoMicCheckbox, pttRow, deviceRow,
                                         levelRow, auxHeader, auxCheckbox, auxDeviceRow, auxGainRow,
                                         notificationsHeader, soundsCheckbox, volumeRow,
                                         speechCheckbox, selfTalkCheckbox, pttSoundCheckbox])
@@ -330,6 +341,7 @@ final class SettingsWindowController: NSWindowController, NSWindowDelegate {
         inputGainSlider.doubleValue = Double(mc.inputGain * 100)
         updateInputGainLabel()
         nrCheckbox.state = mc.inputNoiseReduction ? .on : .off
+        stereoMicCheckbox.state = mc.stereoMic ? .on : .off
 
         auxCheckbox.state = mc.auxEnabled ? .on : .off
         auxGainSlider.doubleValue = Double(mc.auxGain * 100)
@@ -390,6 +402,16 @@ final class SettingsWindowController: NSWindowController, NSWindowDelegate {
         }
     }
 
+    @objc private func stereoMicChanged() {
+        let on = stereoMicCheckbox.state == .on
+        mainController?.stereoMic = on
+        // Channel count only takes effect when the capture device (re)starts, so restart it live.
+        if let mc = mainController, mc.micStreamId != 0 {
+            client.setCaptureChannels(streamId: mc.micStreamId, channels: on ? 2 : 1)
+            client.audioRestart()
+        }
+    }
+
     private func updateInputGainLabel() {
         let pct = Int(inputGainSlider.doubleValue.rounded())
         inputGainValueLabel.stringValue = "\(pct)%"

clients/windows/VoiceCat.App/Forms/AudioSettingsForm.cs

@@ -28,6 +28,7 @@ public sealed class AudioSettingsForm : Form
     private readonly int _origVadSlider;
     private readonly int _origMicGain;
     private readonly bool _origMicNoiseReduction;
+    private readonly bool _origStereoMic;
     private readonly Keys _origPttKey;
     private readonly bool _origAuxEnabled;
     private readonly string? _origAuxDeviceId;
@@ -44,6 +45,7 @@ public sealed class AudioSettingsForm : Form
     private readonly TrackBar _trkVad;
     private readonly TrackBar _trkGain;
     private readonly CheckBox _chkNoiseReduction;
+    private readonly CheckBox _chkStereoMic;
     private readonly CheckBox _chkAux;
     private readonly Label _lblAuxDevice;
     private readonly ComboBox _cboAuxDevice;
@@ -70,6 +72,7 @@ public sealed class AudioSettingsForm : Form
         _origVadSlider = settings.VadThresholdSlider;
         _origMicGain = settings.MicGain;
         _origMicNoiseReduction = settings.MicNoiseReduction;
+        _origStereoMic = settings.StereoMic;
         _origPttKey = _pttKey;
         _origAuxEnabled = settings.AuxEnabled;
         _origAuxDeviceId = settings.AuxDeviceId;
@@ -81,7 +84,7 @@ public sealed class AudioSettingsForm : Form
         MinimizeBox = false;
         StartPosition = FormStartPosition.CenterParent;
         AutoScaleMode = AutoScaleMode.Font;
-        ClientSize = new Size(420, 583);
+        ClientSize = new Size(420, 611);
 
         // ── Device row ────────────────────────────────────────────────────────
         var lblDevice = new Label
@@ -217,6 +220,23 @@ public sealed class AudioSettingsForm : Form
         };
         _chkNoiseReduction.CheckedChanged += ChkNoiseReduction_CheckedChanged;
 
+        // ── Stereo microphone ─────────────────────────────────────────────────
+        // Opens the mic capture device 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. Toggling while connected restarts the capture device (vc_audio_restart) so the
+        // new channel count takes effect immediately.
+        _chkStereoMic = new CheckBox
+        {
+            Text = "&Stereo microphone",
+            Location = new Point(12, 364),
+            AutoSize = true,
+            Checked = settings.StereoMic,
+            AccessibleName = "Stereo microphone",
+            AccessibleDescription =
+                "Capture your microphone in stereo. Only transmitted in stereo on a stereo channel.",
+        };
+        _chkStereoMic.CheckedChanged += ChkStereoMic_CheckedChanged;
+
         // ── Aux input stream ──────────────────────────────────────────────────
         // A second outgoing stream from another hardware input device (e.g. line-in / aux),
         // captured client-side and fed to the core. Device + volume only — aux is always-on
@@ -224,7 +244,7 @@ public sealed class AudioSettingsForm : Form
         _chkAux = new CheckBox
         {
             Text = "&Aux stream (second input device)",
-            Location = new Point(12, 376),
+            Location = new Point(12, 404),
             AutoSize = true,
             Checked = settings.AuxEnabled,
             AccessibleName = "Enable aux input stream",
@@ -236,12 +256,12 @@ public sealed class AudioSettingsForm : Form
         _lblAuxDevice = new Label
         {
             Text = "Aux d&evice:",
-            Location = new Point(12, 406),
+            Location = new Point(12, 434),
             AutoSize = true,
         };
         _cboAuxDevice = new ComboBox
         {
-            Location = new Point(12, 426),
+            Location = new Point(12, 454),
             Width = 300,
             DropDownStyle = ComboBoxStyle.DropDownList,
             DisplayMember = "Name",
@@ -254,7 +274,7 @@ public sealed class AudioSettingsForm : Form
         _btnAuxRefresh = new Button
         {
             Text = "Re&fresh",
-            Location = new Point(320, 424),
+            Location = new Point(320, 452),
             Size = new Size(80, 26),
         };
         _btnAuxRefresh.Click += (_, _) => LoadAuxDevices();
@@ -262,12 +282,12 @@ public sealed class AudioSettingsForm : Form
         _lblAuxGain = new Label
         {
             Text = "Aux vo&lume:",
-            Location = new Point(12, 462),
+            Location = new Point(12, 490),
             AutoSize = true,
         };
         _trkAuxGain = new TrackBar
         {
-            Location = new Point(12, 482),
+            Location = new Point(12, 510),
             Size = new Size(200, 45),
             Minimum = 0,
             Maximum = 300,
@@ -286,14 +306,14 @@ public sealed class AudioSettingsForm : Form
         {
             Text = "&OK",
             DialogResult = DialogResult.OK,
-            Location = new Point(228, 544),
+            Location = new Point(228, 572),
             Size = new Size(80, 27),
         };
         var btnCancel = new Button
         {
             Text = "&Cancel",
             DialogResult = DialogResult.Cancel,
-            Location = new Point(316, 544),
+            Location = new Point(316, 572),
             Size = new Size(80, 27),
         };
 
@@ -310,7 +330,7 @@ public sealed class AudioSettingsForm : Form
             lblDevice, _cboDevice, _btnRefresh,
             lblMode, _radioVad, _lblSensitivity, _trkVad,
             _radioPtt, _lblPttKey, _btnChangePtt, _radioAlwaysOn,
-            lblGain, _trkGain, _chkNoiseReduction,
+            lblGain, _trkGain, _chkNoiseReduction, _chkStereoMic,
             _chkAux, _lblAuxDevice, _cboAuxDevice, _btnAuxRefresh, _lblAuxGain, _trkAuxGain,
             btnOk, btnCancel,
         ]);
@@ -455,6 +475,17 @@ public sealed class AudioSettingsForm : Form
             _client.SetInputNoiseReduction(_chkNoiseReduction.Checked);
     }
 
+    private void ChkStereoMic_CheckedChanged(object? sender, EventArgs e)
+    {
+        _settings.StereoMic = _chkStereoMic.Checked;
+        // Channel count only takes effect when the capture device (re)starts, so restart it live.
+        if (_micStreamId != 0)
+        {
+            _client.SetCaptureChannels(_micStreamId, _chkStereoMic.Checked ? 2u : 1u);
+            _client.AudioRestart();
+        }
+    }
+
     private void BtnChangePtt_Click(object? sender, EventArgs e)
     {
         using var dlg = new PttKeyCaptureDialog(_pttKey);
@@ -476,6 +507,7 @@ public sealed class AudioSettingsForm : Form
         _settings.VadThresholdSlider = _origVadSlider;
         _settings.MicGain = _origMicGain;
         _settings.MicNoiseReduction = _origMicNoiseReduction;
+        _settings.StereoMic = _origStereoMic;
         _settings.PttKey = (int)_origPttKey;
 
         if (_micStreamId != 0)
@@ -486,6 +518,12 @@ public sealed class AudioSettingsForm : Form
                 _client.SetVadThreshold(0.1f * (1f - (_origVadSlider - 1f) / 99f));
             _client.SetInputGain(_origMicGain / 100f);
             _client.SetInputNoiseReduction(_origMicNoiseReduction);
+            // Restore capture channel count; restart the device only if it actually changed.
+            if (_origStereoMic != _chkStereoMic.Checked)
+            {
+                _client.SetCaptureChannels(_micStreamId, _origStereoMic ? 2u : 1u);
+                _client.AudioRestart();
+            }
         }
 
         // Aux: restore originals to settings and re-apply live (order: device + gain first so a

clients/windows/VoiceCat.App/Forms/MainForm.cs

@@ -646,6 +646,7 @@ public partial class MainForm : Form
                 var mode = (VcInputMode)_voiceSettings.InputMode;
                 if (_voiceSettings.InputDeviceId is string devId)
                     _client.SetInputDevice(streamId, devId);
+                _client.SetCaptureChannels(streamId, _voiceSettings.StereoMic ? 2u : 1u);
                 _client.SetInputMode(mode);
                 if (mode == VcInputMode.VoiceActivation)
                     _client.SetVadThreshold(VadThresholdFromSettings());

clients/windows/VoiceCat.App/Models/VoiceSettings.cs

@@ -26,6 +26,11 @@ public sealed class VoiceSettings
     /// cleaned signal. Independent of the per-listener receive-side NR.</summary>
     public bool MicNoiseReduction { get; set; } = false;
 
+    /// <summary>Capture the mic in stereo (interleaved L/R) instead of mono. Off by default. Real
+    /// stereo only reaches the wire on a stereo channel; on a mono channel the core folds the mic
+    /// to mono. Applied when the capture device next starts (Join Voice or an audio restart).</summary>
+    public bool StereoMic { get; set; } = false;
+
     /// <summary>Push-to-talk key, stored as the integer value of System.Windows.Forms.Keys.</summary>
     public int PttKey { get; set; } = (int)Keys.F8;
 

clients/windows/VoiceCat.Interop/NativeMethods.cs

@@ -65,6 +65,12 @@ internal static partial class NativeMethods
     [LibraryImport(LibName, StringMarshalling = StringMarshalling.Utf8)]
     internal static partial VcResult vc_set_input_device(nint c, uint streamId, string? deviceId);
 
+    [LibraryImport(LibName)]
+    internal static partial VcResult vc_set_capture_channels(nint c, uint streamId, uint channels);
+
+    [LibraryImport(LibName)]
+    internal static partial VcResult vc_audio_restart(nint c);
+
     [LibraryImport(LibName)]
     internal static partial VcResult vc_set_input_mode(nint c, VcInputMode mode);
 

clients/windows/VoiceCat.Interop/VoiceCatClient.cs

@@ -212,6 +212,19 @@ public sealed class VoiceCatClient : IDisposable
     public VcResult SetInputDevice(uint streamId, string? deviceId) =>
         NativeMethods.vc_set_input_device(_handle.DangerousGetHandle(), streamId, deviceId);
 
+    /// <summary>Sets the mic capture channel count (1 = mono, 2 = stereo) for the given stream.
+    /// Applied when the capture device next (re)starts — call before Join Voice, or pair with an
+    /// audio restart to take effect live. Real stereo only reaches the wire on a stereo channel;
+    /// the core folds a stereo mic to mono on a mono channel.</summary>
+    public VcResult SetCaptureChannels(uint streamId, uint channels) =>
+        NativeMethods.vc_set_capture_channels(_handle.DangerousGetHandle(), streamId, channels);
+
+    /// <summary>Uninitializes and re-initializes the capture and playback devices on a running
+    /// engine, applying pending changes (e.g. capture channel count) that only take effect on a
+    /// device restart. No-op if audio isn't running.</summary>
+    public VcResult AudioRestart() =>
+        NativeMethods.vc_audio_restart(_handle.DangerousGetHandle());
+
     public VcResult SetInputMode(VcInputMode mode) =>
         NativeMethods.vc_set_input_mode(_handle.DangerousGetHandle(), mode);