Bump version: 0.1.1 → 0.1.2

- Changed the port for the development server from 8000 to 17493 to avoid conflicts and improve accessibility during local development.

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.1.1
+current_version = 0.1.2
 commit = True
 tag = True
 tag_name = v{new_version}

app/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@voicebox/app",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "private": true,
   "type": "module",
   "scripts": {

app/src/components/ServerSettings/UpdateStatus.tsx

@@ -71,11 +71,23 @@ export function UpdateStatus() {
 
         {status.downloading && (
           <div className="space-y-2">
-            <div className="flex items-center gap-2 text-sm">
+            <div className="flex items-center justify-between text-sm">
-              <Download className="h-4 w-4" />
+              <div className="flex items-center gap-2">
-              Downloading update...
+                <Download className="h-4 w-4" />
+                Downloading update...
+              </div>
+              {status.downloadProgress !== undefined && (
+                <span className="text-muted-foreground">
+                  {status.downloadProgress}%
+                </span>
+              )}
             </div>
-            <Progress />
+            <Progress value={status.downloadProgress} />
+            {status.downloadedBytes !== undefined && status.totalBytes !== undefined && status.totalBytes > 0 && (
+              <div className="text-xs text-muted-foreground">
+                {(status.downloadedBytes / 1024 / 1024).toFixed(1)} MB / {(status.totalBytes / 1024 / 1024).toFixed(1)} MB
+              </div>
+            )}
           </div>
         )}
 

app/src/components/VoiceProfiles/AudioSampleRecording.tsx

@@ -1,4 +1,4 @@
-import { Mic, Square, Play, Pause } from 'lucide-react';
+import { Mic, Pause, Play, Square } from 'lucide-react';
 import { Button } from '@/components/ui/button';
 import { FormControl, FormItem, FormLabel, FormMessage } from '@/components/ui/form';
 import { formatAudioDuration } from '@/lib/utils/audio';
@@ -35,12 +35,7 @@ export function AudioSampleRecording({
         <div className="space-y-4">
           {!isRecording && !file && (
             <div className="flex flex-col items-center justify-center gap-4 p-4 border-2 border-dashed rounded-lg min-h-[180px]">
-              <Button
+              <Button type="button" onClick={onStart} size="lg" className="flex items-center gap-2">
-                type="button"
-                onClick={onStart}
-                size="lg"
-                className="flex items-center gap-2"
-              >
                 <Mic className="h-5 w-5" />
                 Start Recording
               </Button>
@@ -81,16 +76,9 @@ export function AudioSampleRecording({
                 <Mic className="h-5 w-5 text-primary" />
                 <span className="font-medium">Recording complete</span>
               </div>
-              <p className="text-sm text-muted-foreground text-center">
+              <p className="text-sm text-muted-foreground text-center">File: {file.name}</p>
-                File: {file.name}
-              </p>
               <div className="flex gap-2">
-                <Button
+                <Button type="button" size="icon" variant="outline" onClick={onPlayPause}>
-                  type="button"
-                  size="icon"
-                  variant="outline"
-                  onClick={onPlayPause}
-                >
                   {isPlaying ? <Pause className="h-4 w-4" /> : <Play className="h-4 w-4" />}
                 </Button>
                 <Button

app/src/components/VoiceProfiles/AudioSampleSystem.tsx

@@ -1,4 +1,4 @@
-import { Monitor, Square, Play, Pause, Mic } from 'lucide-react';
+import { Mic, Monitor, Pause, Play, Square } from 'lucide-react';
 import { Button } from '@/components/ui/button';
 import { FormControl, FormItem, FormLabel, FormMessage } from '@/components/ui/form';
 import { formatAudioDuration } from '@/lib/utils/audio';
@@ -35,12 +35,7 @@ export function AudioSampleSystem({
         <div className="space-y-4">
           {!isRecording && !file && (
             <div className="flex flex-col items-center justify-center gap-4 p-4 border-2 border-dashed rounded-lg min-h-[180px]">
-              <Button
+              <Button type="button" onClick={onStart} size="lg" className="flex items-center gap-2">
-                type="button"
-                onClick={onStart}
-                size="lg"
-                className="flex items-center gap-2"
-              >
                 <Monitor className="h-5 w-5" />
                 Start Capture
               </Button>
@@ -81,16 +76,9 @@ export function AudioSampleSystem({
                 <Monitor className="h-5 w-5 text-primary" />
                 <span className="font-medium">Capture complete</span>
               </div>
-              <p className="text-sm text-muted-foreground text-center">
+              <p className="text-sm text-muted-foreground text-center">File: {file.name}</p>
-                File: {file.name}
-              </p>
               <div className="flex gap-2">
-                <Button
+                <Button type="button" size="icon" variant="outline" onClick={onPlayPause}>
-                  type="button"
-                  size="icon"
-                  variant="outline"
-                  onClick={onPlayPause}
-                >
                   {isPlaying ? <Pause className="h-4 w-4" /> : <Play className="h-4 w-4" />}
                 </Button>
                 <Button

app/src/components/VoiceProfiles/SampleUpload.tsx

@@ -22,15 +22,15 @@ import {
 import { Tabs, TabsContent, TabsList, TabsTrigger } from '@/components/ui/tabs';
 import { Textarea } from '@/components/ui/textarea';
 import { useToast } from '@/components/ui/use-toast';
-import { useAudioRecording } from '@/lib/hooks/useAudioRecording';
 import { useAudioPlayer } from '@/lib/hooks/useAudioPlayer';
+import { useAudioRecording } from '@/lib/hooks/useAudioRecording';
 import { useAddSample, useProfile } from '@/lib/hooks/useProfiles';
 import { useSystemAudioCapture } from '@/lib/hooks/useSystemAudioCapture';
 import { useTranscription } from '@/lib/hooks/useTranscription';
 import { isTauri } from '@/lib/tauri';
-import { AudioSampleUpload } from './AudioSampleUpload';
 import { AudioSampleRecording } from './AudioSampleRecording';
 import { AudioSampleSystem } from './AudioSampleSystem';
+import { AudioSampleUpload } from './AudioSampleUpload';
 
 const sampleSchema = z.object({
   file: z.instanceof(File, { message: 'Please select an audio file' }),

app/src/hooks/useAutoUpdater.ts

@@ -1,4 +1,4 @@
-import { useEffect, useState } from 'react';
+import { useCallback, useEffect, useState } from 'react';
 import { check, type Update } from '@tauri-apps/plugin-updater';
 import { relaunch } from '@tauri-apps/plugin-process';
 
@@ -9,6 +9,9 @@ export interface UpdateStatus {
   downloading: boolean;
   installing: boolean;
   error?: string;
+  downloadProgress?: number; // 0-100 percentage
+  downloadedBytes?: number;
+  totalBytes?: number;
 }
 
 const isTauri = () => {
@@ -25,7 +28,7 @@ export function useAutoUpdater(checkOnMount = false) {
 
   const [update, setUpdate] = useState<Update | null>(null);
 
-  const checkForUpdates = async () => {
+  const checkForUpdates = useCallback(async () => {
     if (!isTauri()) {
       return;
     }
@@ -61,7 +64,7 @@ export function useAutoUpdater(checkOnMount = false) {
         error: error instanceof Error ? error.message : 'Failed to check for updates',
       });
     }
-  };
+  }, []);
 
   const downloadAndInstall = async () => {
     if (!update || !isTauri()) return;
@@ -69,19 +72,40 @@ export function useAutoUpdater(checkOnMount = false) {
     try {
       setStatus((prev) => ({ ...prev, downloading: true, error: undefined }));
 
+      let downloadedBytes = 0;
+      let totalBytes = 0;
+
       await update.downloadAndInstall((event) => {
         switch (event.event) {
           case 'Started':
-            setStatus((prev) => ({ ...prev, downloading: true }));
+            totalBytes = event.data.contentLength || 0;
+            downloadedBytes = 0;
+            setStatus((prev) => ({ 
+              ...prev, 
+              downloading: true,
+              totalBytes,
+              downloadedBytes: 0,
+              downloadProgress: 0
+            }));
             break;
-          case 'Progress':
+          case 'Progress': {
-            console.log(`Downloaded ${event.data.chunkLength} bytes`);
+            downloadedBytes += event.data.chunkLength;
+            const progress = totalBytes > 0 
+              ? Math.round((downloadedBytes / totalBytes) * 100)
+              : undefined;
+            setStatus((prev) => ({
+              ...prev,
+              downloadedBytes,
+              downloadProgress: progress
+            }));
             break;
+          }
           case 'Finished':
             setStatus((prev) => ({
               ...prev,
               downloading: false,
               installing: true,
+              downloadProgress: 100
             }));
             break;
         }
@@ -93,6 +117,9 @@ export function useAutoUpdater(checkOnMount = false) {
         ...prev,
         downloading: false,
         installing: false,
+        downloadProgress: undefined,
+        downloadedBytes: undefined,
+        totalBytes: undefined,
         error: error instanceof Error ? error.message : 'Failed to install update',
       }));
     }
@@ -102,7 +129,7 @@ export function useAutoUpdater(checkOnMount = false) {
     if (checkOnMount && isTauri()) {
       checkForUpdates();
     }
-  }, [checkOnMount]);
+  }, [checkOnMount, checkForUpdates]);
 
   return {
     status,

app/src/lib/hooks/useAudioRecording.ts

@@ -1,5 +1,6 @@
-import { useState, useRef, useCallback, useEffect } from 'react';
+import { useCallback, useEffect, useRef, useState } from 'react';
 import { isTauri } from '@/lib/tauri';
+import { convertToWav } from '@/lib/utils/audio';
 
 interface UseAudioRecordingOptions {
   maxDurationSeconds?: number;
@@ -85,13 +86,26 @@ export function useAudioRecording({
         }
       };
 
-      mediaRecorder.onstop = () => {
+      mediaRecorder.onstop = async () => {
-        const blob = new Blob(chunksRef.current, { type: 'audio/webm' });
+        const webmBlob = new Blob(chunksRef.current, { type: 'audio/webm' });
-        // Pass the actual recorded duration
+
-        const recordedDuration = startTimeRef.current 
+        // Convert to WAV format to avoid needing ffmpeg on backend
-          ? (Date.now() - startTimeRef.current) / 1000 
+        try {
-          : undefined;
+          const wavBlob = await convertToWav(webmBlob);
-        onRecordingComplete?.(blob, recordedDuration);
+
+          // Pass the actual recorded duration
+          const recordedDuration = startTimeRef.current
+            ? (Date.now() - startTimeRef.current) / 1000
+            : undefined;
+          onRecordingComplete?.(wavBlob, recordedDuration);
+        } catch (err) {
+          console.error('Error converting audio to WAV:', err);
+          // Fallback to original blob if conversion fails
+          const recordedDuration = startTimeRef.current
+            ? (Date.now() - startTimeRef.current) / 1000
+            : undefined;
+          onRecordingComplete?.(webmBlob, recordedDuration);
+        }
 
         // Stop all tracks
         streamRef.current?.getTracks().forEach((track) => {

app/src/lib/hooks/useSystemAudioCapture.ts

@@ -22,6 +22,7 @@ export function useSystemAudioCapture({
   const timerRef = useRef<number | null>(null);
   const startTimeRef = useRef<number | null>(null);
   const stopRecordingRef = useRef<(() => Promise<void>) | null>(null);
+  const isRecordingRef = useRef(false);
 
   // Check if system audio capture is supported
   useEffect(() => {
@@ -62,6 +63,7 @@ export function useSystemAudioCapture({
       });
 
       setIsRecording(true);
+      isRecordingRef.current = true;
       startTimeRef.current = Date.now();
 
       // Start timer
@@ -93,6 +95,7 @@ export function useSystemAudioCapture({
 
     try {
       setIsRecording(false);
+      isRecordingRef.current = false;
 
       if (timerRef.current !== null) {
         clearInterval(timerRef.current);
@@ -130,32 +133,37 @@ export function useSystemAudioCapture({
   }, [stopRecording]);
 
   const cancelRecording = useCallback(async () => {
-    if (isRecording) {
+    if (isRecordingRef.current) {
       await stopRecording();
     }
 
     setIsRecording(false);
+    isRecordingRef.current = false;
     setDuration(0);
 
     if (timerRef.current !== null) {
       clearInterval(timerRef.current);
       timerRef.current = null;
     }
-  }, [isRecording, stopRecording]);
+  }, [stopRecording]);
 
-  // Cleanup on unmount
+  // Cleanup on unmount only
   useEffect(() => {
     return () => {
       if (timerRef.current !== null) {
         clearInterval(timerRef.current);
+        timerRef.current = null;
       }
       // Cancel recording on unmount if still recording
-      if (isRecording) {
+      if (isRecordingRef.current && isTauri()) {
-        void cancelRecording();
+        // Call stop directly without the callback to avoid stale closure
+        invoke('stop_system_audio_capture').catch((err) => {
+          console.error('Error stopping audio capture on unmount:', err);
+        });
       }
     };
-    // biome-ignore lint/correctness/useExhaustiveDependencies: cancelRecording is stable
+    // biome-ignore lint/correctness/useExhaustiveDependencies: Only run on unmount
-  }, [isRecording]);
+  }, []);
 
   return {
     isRecording,

app/src/lib/utils/audio.ts

@@ -16,3 +16,104 @@ export function formatAudioDuration(seconds: number): string {
   const secs = Math.floor(seconds % 60);
   return `${mins}:${secs.toString().padStart(2, '0')}`;
 }
+
+/**
+ * Convert any audio blob to WAV format using Web Audio API.
+ * This ensures compatibility without requiring ffmpeg on the backend.
+ */
+export async function convertToWav(audioBlob: Blob): Promise<Blob> {
+  // Create audio context
+  const audioContext = new AudioContext();
+
+  // Read blob as array buffer
+  const arrayBuffer = await audioBlob.arrayBuffer();
+
+  // Decode audio data
+  const audioBuffer = await audioContext.decodeAudioData(arrayBuffer);
+
+  // Convert to WAV
+  const wavBlob = audioBufferToWav(audioBuffer);
+
+  // Close audio context to free resources
+  await audioContext.close();
+
+  return wavBlob;
+}
+
+/**
+ * Convert AudioBuffer to WAV blob.
+ */
+function audioBufferToWav(buffer: AudioBuffer): Blob {
+  const numberOfChannels = buffer.numberOfChannels;
+  const sampleRate = buffer.sampleRate;
+  const format = 1; // PCM
+  const bitDepth = 16;
+
+  const bytesPerSample = bitDepth / 8;
+  const blockAlign = numberOfChannels * bytesPerSample;
+
+  // Interleave channels
+  const interleaved = interleaveChannels(buffer);
+
+  // Create WAV file
+  const dataLength = interleaved.length * bytesPerSample;
+  const buffer2 = new ArrayBuffer(44 + dataLength);
+  const view = new DataView(buffer2);
+
+  // Write WAV header
+  writeString(view, 0, 'RIFF');
+  view.setUint32(4, 36 + dataLength, true);
+  writeString(view, 8, 'WAVE');
+  writeString(view, 12, 'fmt ');
+  view.setUint32(16, 16, true); // fmt chunk size
+  view.setUint16(20, format, true); // audio format (PCM)
+  view.setUint16(22, numberOfChannels, true);
+  view.setUint32(24, sampleRate, true);
+  view.setUint32(28, sampleRate * blockAlign, true); // byte rate
+  view.setUint16(32, blockAlign, true);
+  view.setUint16(34, bitDepth, true);
+  writeString(view, 36, 'data');
+  view.setUint32(40, dataLength, true);
+
+  // Write audio data
+  floatTo16BitPCM(view, 44, interleaved);
+
+  return new Blob([buffer2], { type: 'audio/wav' });
+}
+
+/**
+ * Interleave multiple channels into a single array.
+ */
+function interleaveChannels(buffer: AudioBuffer): Float32Array {
+  const numberOfChannels = buffer.numberOfChannels;
+  const length = buffer.length;
+  const interleaved = new Float32Array(length * numberOfChannels);
+
+  for (let channel = 0; channel < numberOfChannels; channel++) {
+    const channelData = buffer.getChannelData(channel);
+    for (let i = 0; i < length; i++) {
+      interleaved[i * numberOfChannels + channel] = channelData[i];
+    }
+  }
+
+  return interleaved;
+}
+
+/**
+ * Write string to DataView.
+ */
+function writeString(view: DataView, offset: number, string: string): void {
+  for (let i = 0; i < string.length; i++) {
+    view.setUint8(offset + i, string.charCodeAt(i));
+  }
+}
+
+/**
+ * Convert float32 audio data to 16-bit PCM.
+ */
+function floatTo16BitPCM(view: DataView, offset: number, input: Float32Array): void {
+  for (let i = 0; i < input.length; i++, offset += 2) {
+    const s = Math.max(-1, Math.min(1, input[i]));
+    view.setInt16(offset, s < 0 ? s * 0x8000 : s * 0x7fff, true);
+  }
+}

backend/main.py

@@ -47,7 +47,7 @@ app.add_middleware(
 @app.get("/")
 async def root():
     """Root endpoint."""
-    return {"message": "voicebox API", "version": "0.1.1"}
+    return {"message": "voicebox API", "version": "0.1.2"}
 
 
 @app.get("/health", response_model=models.HealthResponse)

backend/voicebox-server.spec

@@ -4,15 +4,16 @@ from PyInstaller.utils.hooks import collect_submodules
 from PyInstaller.utils.hooks import copy_metadata
 
 datas = []
-hiddenimports = ['backend', 'backend.main', 'backend.config', 'backend.database', 'backend.models', 'backend.profiles', 'backend.history', 'backend.tts', 'backend.transcribe', 'backend.utils.audio', 'backend.utils.cache', 'backend.utils.progress', 'backend.utils.hf_progress', 'backend.utils.validation', 'torch', 'transformers', 'fastapi', 'uvicorn', 'sqlalchemy', 'librosa', 'soundfile', 'qwen_tts', 'qwen_tts.inference', 'qwen_tts.inference.qwen3_tts_model', 'qwen_tts.inference.qwen3_tts_tokenizer', 'qwen_tts.core', 'qwen_tts.cli']
+hiddenimports = ['backend', 'backend.main', 'backend.config', 'backend.database', 'backend.models', 'backend.profiles', 'backend.history', 'backend.tts', 'backend.transcribe', 'backend.utils.audio', 'backend.utils.cache', 'backend.utils.progress', 'backend.utils.hf_progress', 'backend.utils.validation', 'torch', 'transformers', 'fastapi', 'uvicorn', 'sqlalchemy', 'librosa', 'soundfile', 'qwen_tts', 'qwen_tts.inference', 'qwen_tts.inference.qwen3_tts_model', 'qwen_tts.inference.qwen3_tts_tokenizer', 'qwen_tts.core', 'qwen_tts.cli', 'pkg_resources.extern']
 datas += collect_data_files('qwen_tts')
 datas += copy_metadata('qwen-tts')
 hiddenimports += collect_submodules('qwen_tts')
+hiddenimports += collect_submodules('jaraco')
 
 
 a = Analysis(
     ['server.py'],
-    pathex=['/Users/jamespine/Projects/voice/Qwen3-TTS'],
+    pathex=['C:\\Users\\ijame\\Projects\\voice\\Qwen3-TTS'],
     binaries=[],
     datas=datas,
     hiddenimports=hiddenimports,

landing/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@voicebox/landing",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "description": "Landing page for voicebox.sh",
   "scripts": {
     "dev": "bun --bun next dev --turbo",

package.json

@@ -1,6 +1,6 @@
 {
   "name": "voicebox",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "private": true,
   "workspaces": [
     "app",
@@ -12,7 +12,7 @@
     "dev": "cd tauri && bun run tauri dev",
     "dev:web": "cd web && bun run dev",
     "dev:landing": "cd landing && bun run dev",
-    "dev:server": "source backend/venv/bin/activate && uvicorn backend.main:app --reload --port 8000",
+    "dev:server": "uvicorn backend.main:app --reload --port 17493",
     "build": "cd tauri && bun run tauri build",
     "build:web": "cd web && bun run build",
     "build:landing": "cd landing && bun run build",

tauri/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@voicebox/tauri",
   "private": true,
-  "version": "0.1.1",
+  "version": "0.1.2",
   "type": "module",
   "scripts": {
     "dev": "vite",

tauri/src-tauri/Cargo.lock

@@ -4490,6 +4490,7 @@ dependencies = [
  "coreaudio-sys",
  "hound",
  "objc",
+ "scopeguard",
  "screencapturekit",
  "serde",
  "serde_json",

tauri/src-tauri/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "voicebox"
-version = "0.1.1"
+version = "0.1.2"
 description = "A production-quality desktop app for Qwen3-TTS voice cloning and generation"
 authors = ["you"]
 license = ""
@@ -22,6 +22,7 @@ serde_json = "1.0"
 tokio = { version = "1", features = ["full"] }
 hound = "3.5"
 base64 = "0.22"
+scopeguard = "1.2.0"
 
 [target.'cfg(target_os = "macos")'.dependencies]
 screencapturekit = { version = "1", features = ["async"] }
@@ -31,7 +32,7 @@ core-foundation-sys = "0.8"
 
 [target.'cfg(target_os = "windows")'.dependencies]
 wasapi = "0.22"
-windows = { version = "0.62", features = ["Win32_Foundation", "Win32_UI_WindowsAndMessaging"] }
+windows = { version = "0.62", features = ["Win32_Foundation", "Win32_UI_WindowsAndMessaging", "Win32_System_Com"] }
 
 [target.'cfg(not(any(target_os = "android", target_os = "ios")))'.dependencies]
 tauri-plugin-updater = "2.0"

tauri/src-tauri/Entitlements.plist

@@ -8,5 +8,7 @@
     <true/>
     <key>com.apple.security.cs.disable-library-validation</key>
     <true/>
+    <key>com.apple.security.device.audio-input</key>
+    <true/>
 </dict>
 </plist>

tauri/src-tauri/src/audio_capture/macos.rs

@@ -163,25 +163,67 @@ fn extract_audio_samples(sample_buffer: CMSampleBuffer) -> Result<Vec<f32>, Stri
         .audio_buffer_list()
         .ok_or_else(|| "Failed to get audio buffer list".to_string())?;
 
-    let mut samples = Vec::new();
+    let buffers: Vec<_> = audio_buffer_list.iter().collect();
+    let num_buffers = buffers.len();
     
-    // Iterate through audio buffers
+    if num_buffers == 0 {
-    for buffer in audio_buffer_list.iter() {
+        return Ok(Vec::new());
-        // Get raw bytes and interpret as f32 samples
+    }
+
+    // ScreenCaptureKit on macOS provides audio in Float32 format
+    // The audio can be either:
+    // - Interleaved (1 buffer with L,R,L,R,... samples)
+    // - Planar (2 buffers, one for L channel, one for R channel)
+    
+    if num_buffers == 1 {
+        // Interleaved stereo or mono in a single buffer
+        let buffer = &buffers[0];
         let data_bytes = buffer.data();
         let num_samples = data_bytes.len() / std::mem::size_of::<f32>();
         
         if num_samples > 0 {
             unsafe {
-                // Interpret bytes as f32 samples
                 let data_ptr = data_bytes.as_ptr() as *const f32;
                 let data = std::slice::from_raw_parts(data_ptr, num_samples);
-                samples.extend_from_slice(data);
+                return Ok(data.to_vec());
             }
         }
+    } else {
+        // Planar format - separate buffer for each channel
+        // We need to interleave them: L0, R0, L1, R1, ...
+        let mut channel_data: Vec<Vec<f32>> = Vec::new();
+        let mut max_samples = 0;
+        
+        for buffer in &buffers {
+            let data_bytes = buffer.data();
+            let num_samples = data_bytes.len() / std::mem::size_of::<f32>();
+            
+            if num_samples > 0 {
+                unsafe {
+                    let data_ptr = data_bytes.as_ptr() as *const f32;
+                    let data = std::slice::from_raw_parts(data_ptr, num_samples);
+                    channel_data.push(data.to_vec());
+                    max_samples = max_samples.max(num_samples);
+                }
+            }
+        }
+        
+        // Interleave the channels
+        let mut interleaved = Vec::with_capacity(max_samples * num_buffers);
+        for i in 0..max_samples {
+            for channel in &channel_data {
+                if i < channel.len() {
+                    interleaved.push(channel[i]);
+                } else {
+                    interleaved.push(0.0); // Pad with silence if needed
+                }
+            }
+        }
+        
+        return Ok(interleaved);
     }
 
-    Ok(samples)
+    Ok(Vec::new())
 }
 
 fn samples_to_wav(samples: &[f32], sample_rate: u32, channels: u16) -> Result<Vec<u8>, String> {

tauri/src-tauri/src/audio_capture/mod.rs

@@ -18,6 +18,7 @@ pub struct AudioCaptureState {
     pub sample_rate: Arc<Mutex<u32>>,
     pub channels: Arc<Mutex<u16>>,
     pub stop_tx: Arc<Mutex<Option<tokio::sync::mpsc::Sender<()>>>>,
+    pub error: Arc<Mutex<Option<String>>>,
     #[cfg(target_os = "macos")]
     pub stream: Arc<Mutex<Option<SCStream>>>,
 }
@@ -29,6 +30,7 @@ impl AudioCaptureState {
             sample_rate: Arc::new(Mutex::new(44100)),
             channels: Arc::new(Mutex::new(2)),
             stop_tx: Arc::new(Mutex::new(None)),
+            error: Arc::new(Mutex::new(None)),
             #[cfg(target_os = "macos")]
             stream: Arc::new(Mutex::new(None)),
         }
@@ -36,5 +38,6 @@ impl AudioCaptureState {
 
     pub fn reset(&self) {
         *self.samples.lock().unwrap() = Vec::new();
+        *self.error.lock().unwrap() = None;
     }
 }

tauri/src-tauri/src/audio_capture/windows.rs

@@ -5,8 +5,8 @@ use std::io::Cursor;
 use std::sync::Arc;
 use std::sync::atomic::{AtomicBool, Ordering};
 use std::thread;
-use std::time::Duration;
 use wasapi::*;
+use windows::Win32::System::Com::{CoInitializeEx, CoUninitialize, COINIT_MULTITHREADED};
 
 pub async fn start_capture(
     state: &AudioCaptureState,
@@ -19,6 +19,7 @@ pub async fn start_capture(
     let sample_rate_arc = state.sample_rate.clone();
     let channels_arc = state.channels.clone();
     let stop_tx = state.stop_tx.clone();
+    let error_arc = state.error.clone();
 
     // Use AtomicBool for stop signal (works with non-Send types)
     let stop_flag = Arc::new(AtomicBool::new(false));
@@ -36,13 +37,29 @@ pub async fn start_capture(
     // Spawn capture task on a dedicated thread (WASAPI COM objects are not Send)
     // All WASAPI objects must be created and used on the same thread
     thread::spawn(move || {
+        // Initialize COM for this thread
+        unsafe {
+            let hr = CoInitializeEx(None, COINIT_MULTITHREADED);
+            if hr.is_err() {
+                eprintln!("Failed to initialize COM: {:?}", hr);
+                return;
+            }
+        }
+
+        // Ensure COM is uninitialized when thread exits
+        let _com_guard = scopeguard::guard((), |_| unsafe {
+            CoUninitialize();
+        });
+
         // Initialize WASAPI on this thread
         let device = match DeviceEnumerator::new()
             .and_then(|enumerator| enumerator.get_default_device(&Direction::Render))
         {
             Ok(d) => d,
             Err(e) => {
-                eprintln!("Failed to get audio device: {}", e);
+                let error_msg = format!("Failed to get audio device: {}", e);
+                eprintln!("{}", error_msg);
+                *error_arc.lock().unwrap() = Some(error_msg);
                 return;
             }
         };
@@ -50,7 +67,9 @@ pub async fn start_capture(
         let mut audio_client = match device.get_iaudioclient() {
             Ok(client) => client,
             Err(e) => {
-                eprintln!("Failed to get audio client: {}", e);
+                let error_msg = format!("Failed to get audio client: {}", e);
+                eprintln!("{}", error_msg);
+                *error_arc.lock().unwrap() = Some(error_msg);
                 return;
             }
         };
@@ -58,7 +77,9 @@ pub async fn start_capture(
         let mix_format = match audio_client.get_mixformat() {
             Ok(format) => format,
             Err(e) => {
-                eprintln!("Failed to get mix format: {}", e);
+                let error_msg = format!("Failed to get mix format: {}", e);
+                eprintln!("{}", error_msg);
+                *error_arc.lock().unwrap() = Some(error_msg);
                 return;
             }
         };
@@ -69,27 +90,53 @@ pub async fn start_capture(
         *sample_rate_arc.lock().unwrap() = mix_format.get_samplespersec();
         *channels_arc.lock().unwrap() = mix_format.get_nchannels();
 
+        // Get device period
+        let (_def_period, min_period) = match audio_client.get_device_period() {
+            Ok(periods) => periods,
+            Err(e) => {
+                eprintln!("Failed to get device period: {}", e);
+                return;
+            }
+        };
+
         // Initialize audio client for loopback with StreamMode
+        // For loopback mode: get Render device, initialize with Capture direction
+        // This triggers AUDCLNT_STREAMFLAGS_LOOPBACK in the wasapi crate
         let stream_mode = StreamMode::EventsShared {
-            autoconvert: false,
+            autoconvert: true,  // Enable automatic format conversion
-            buffer_duration_hns: 0, // 0 = use default buffer size
+            buffer_duration_hns: min_period, // Use minimum period
         };
 
         if let Err(e) = audio_client.initialize_client(&mix_format, &Direction::Capture, &stream_mode) {
-            eprintln!("Failed to initialize audio client: {}", e);
+            let error_msg = format!("Failed to initialize audio client: {}", e);
+            eprintln!("{}", error_msg);
+            *error_arc.lock().unwrap() = Some(error_msg);
             return;
         }
 
+        // Set up event handle for EventsShared mode
+        let h_event = match audio_client.set_get_eventhandle() {
+            Ok(event) => event,
+            Err(e) => {
+                eprintln!("Failed to set event handle: {}", e);
+                return;
+            }
+        };
+
         let capture_client = match audio_client.get_audiocaptureclient() {
             Ok(client) => client,
             Err(e) => {
-                eprintln!("Failed to get capture client: {}", e);
+                let error_msg = format!("Failed to get capture client: {}", e);
+                eprintln!("{}", error_msg);
+                *error_arc.lock().unwrap() = Some(error_msg);
                 return;
             }
         };
 
         if let Err(e) = audio_client.start_stream() {
-            eprintln!("Failed to start stream: {}", e);
+            let error_msg = format!("Failed to start stream: {}", e);
+            eprintln!("{}", error_msg);
+            *error_arc.lock().unwrap() = Some(error_msg);
             return;
         }
 
@@ -145,8 +192,10 @@ pub async fn start_capture(
                 }
             }
 
-            // Sleep briefly to avoid busy-waiting
+            // Wait for event signal (with timeout to allow checking stop flag)
-            thread::sleep(Duration::from_millis(10));
+            if h_event.wait_for_event(100).is_err() {
+                // Timeout is expected - just continue to check stop flag
+            }
         }
 
         // Stop the stream when done
@@ -176,13 +225,18 @@ pub async fn stop_capture(state: &AudioCaptureState) -> Result<String, String> {
     // Wait a bit for capture to stop
     tokio::time::sleep(tokio::time::Duration::from_millis(500)).await;
 
+    // Check if there was an error during capture
+    if let Some(error) = state.error.lock().unwrap().as_ref() {
+        return Err(error.clone());
+    }
+
     // Get samples
     let samples = state.samples.lock().unwrap().clone();
     let sample_rate = *state.sample_rate.lock().unwrap();
     let channels = *state.channels.lock().unwrap();
 
     if samples.is_empty() {
-        return Err("No audio samples captured".to_string());
+        return Err("No audio samples captured. Make sure audio is playing on your system during recording.".to_string());
     }
 
     // Convert to WAV

tauri/src-tauri/src/lib.rs

@@ -0,0 +1 @@
+pub mod audio_capture;

tauri/src-tauri/tauri.conf.json

@@ -1,7 +1,7 @@
 {
   "$schema": "https://schema.tauri.app/config/2",
   "productName": "Voicebox",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "identifier": "sh.voicebox.app",
   "build": {
     "beforeDevCommand": "bun run dev",

tauri/src-tauri/tests/audio_capture_test.rs

@@ -0,0 +1,59 @@
+// NOTE: This test requires system audio to be playing during execution.
+// To run this test successfully:
+//   1. Start playing audio (music, video, etc.)
+//   2. Run: cargo test --test audio_capture_test -- --nocapture
+//   3. The test will capture audio for 5 seconds and verify the output
+
+use voicebox::audio_capture::{AudioCaptureState, start_capture, stop_capture};
+use base64::Engine;
+
+#[tokio::test]
+async fn test_system_audio_capture() {
+    // Create AudioCaptureState
+    let state = AudioCaptureState::new();
+
+    println!("Starting system audio capture with 5 second max duration...");
+
+    // Start capture with 5 second max duration
+    let result = start_capture(&state, 5).await;
+
+    if let Err(e) = result {
+        panic!("Failed to start capture: {}", e);
+    }
+
+    println!("Capture started, waiting 5 seconds...");
+
+    // Wait 5 seconds for capture to complete
+    tokio::time::sleep(tokio::time::Duration::from_secs(5)).await;
+
+    println!("Stopping capture...");
+
+    // Stop capture and get the result
+    let audio_data = stop_capture(&state).await;
+
+    match audio_data {
+        Ok(base64_wav) => {
+            println!("Capture stopped successfully");
+
+            // Validate the returned base64 WAV data
+            println!("Validating base64 WAV data...");
+
+            // Decode base64 to bytes
+            let decoded_bytes = base64::engine::general_purpose::STANDARD
+                .decode(&base64_wav)
+                .expect("Failed to decode base64 data");
+
+            // Verify bytes array is not empty
+            assert!(!decoded_bytes.is_empty(), "Decoded bytes array is empty");
+
+            // Confirm data has content (length > 0)
+            println!("WAV data length: {} bytes", decoded_bytes.len());
+            assert!(decoded_bytes.len() > 0, "WAV data has no content");
+
+            println!("✓ Test passed: Audio capture produced valid WAV data");
+        }
+        Err(e) => {
+            panic!("Failed to stop capture or get audio data: {}", e);
+        }
+    }
+}

web/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@voicebox/web",
   "private": true,
-  "version": "0.1.1",
+  "version": "0.1.2",
   "type": "module",
   "scripts": {
     "dev": "vite",