feat: guided light meditation + Edge TTS neural voice audio

- Add continuous guided light-body-scan meditation (16 segments, 97% coverage) - Edge TTS API endpoint with Xiaoxiao neural voice at -40% rate - Pre-generated chime WAVs and guidance MP3s in public/audio/ - Mute toggle in practice page, sequenced non-overlapping audio - Fix middleware to serve /audio and /api/audio/tts without auth - All LLM output localized to Chinese; fix blank home/settings pages - DeepSeek API compat: json_object, increased max_tokens - Analytics, cron reminders, onboarding flow fixes Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-06-17 20:08:29 +08:00
parent 00db3f6c96
commit 6530b20172
38 changed files with 1371 additions and 182 deletions
@@ -0,0 +1,109 @@
+#!/bin/bash
+# Generate practice audio assets.
+# Prerequisites: Node.js, Python 3, edge-tts (pip install edge-tts)
+set -e
+
+AUDIO_DIR="$(cd "$(dirname "$0")/../public/audio" && pwd)"
+mkdir -p "$AUDIO_DIR"
+
+echo "=== Generating chime sounds (Node.js) ==="
+node -e "
+const fs = require('fs');
+
+function writeWav(filename, buildSamples, sampleRate, durationSec) {
+  const numSamples = Math.floor(sampleRate * durationSec);
+  const samples = new Float32Array(numSamples);
+  buildSamples(samples, sampleRate, durationSec);
+
+  // Convert to 16-bit PCM
+  const buffer = Buffer.alloc(44 + numSamples * 2);
+  buffer.write('RIFF', 0);
+  buffer.writeUInt32LE(36 + numSamples * 2, 4);
+  buffer.write('WAVE', 8);
+  buffer.write('fmt ', 12);
+  buffer.writeUInt32LE(16, 16);       // chunk size
+  buffer.writeUInt16LE(1, 20);        // PCM
+  buffer.writeUInt16LE(1, 22);        // mono
+  buffer.writeUInt32LE(sampleRate, 24);
+  buffer.writeUInt32LE(sampleRate * 2, 28); // byte rate
+  buffer.writeUInt16LE(2, 32);        // block align
+  buffer.writeUInt16LE(16, 34);       // bits per sample
+  buffer.write('data', 36);
+  buffer.writeUInt32LE(numSamples * 2, 40);
+
+  for (let i = 0; i < numSamples; i++) {
+    const s = Math.max(-1, Math.min(1, samples[i]));
+    buffer.writeInt16LE(Math.round(s * 32767), 44 + i * 2);
+  }
+  fs.writeFileSync(filename, buffer);
+  console.log('  Created:', filename);
+}
+
+// Bell-like chime: two sine harmonics (fundamental + octave) with quick decay
+function bellTone(t, freq1, freq2, tStart, duration) {
+  if (t < tStart || t > tStart + duration) return 0;
+  const localT = (t - tStart) / duration;
+  const env = Math.exp(-localT * 6);  // quick exponential decay
+  return 0.35 * env * (
+    Math.sin(2 * Math.PI * freq1 * t) * 0.6 +
+    Math.sin(2 * Math.PI * freq2 * t) * 0.4
+  );
+}
+
+// Start chime — welcoming ascending bell, 2s
+writeWav('$AUDIO_DIR/start-chime.wav', (samples, sr, dur) => {
+  for (let i = 0; i < samples.length; i++) {
+    const t = i / sr;
+    let v = 0;
+    v += bellTone(t, 523.25, 1046.5, 0, 0.9);    // C5
+    v += bellTone(t, 659.25, 1318.5, 0.15, 0.8);   // E5
+    v += bellTone(t, 783.99, 1568.0, 0.3, 0.7);    // G5
+    samples[i] = v;
+  }
+}, 44100, 1.8);
+
+// Soft tick — very short click, 0.08s
+writeWav('$AUDIO_DIR/tick.wav', (samples, sr) => {
+  for (let i = 0; i < samples.length; i++) {
+    const t = i / sr;
+    const env = Math.exp(-t * 80);
+    samples[i] = 0.12 * env * Math.sin(2 * Math.PI * 1200 * t);
+  }
+}, 44100, 0.08);
+
+// End chime — resonant closing chord, 3s
+writeWav('$AUDIO_DIR/end-chime.wav', (samples, sr, dur) => {
+  for (let i = 0; i < samples.length; i++) {
+    const t = i / sr;
+    let v = 0;
+    v += bellTone(t, 261.63, 523.25, 0, 1.5);      // C4 (deeper)
+    v += bellTone(t, 329.63, 659.25, 0.12, 1.3);    // E4
+    v += bellTone(t, 392.00, 783.99, 0.24, 1.1);    // G4
+    v += bellTone(t, 523.25, 1046.5, 0.36, 0.9);    // C5
+    samples[i] = v * 0.9;
+  }
+}, 44100, 3.0);
+
+console.log('Chimes done.');
+"
+
+echo ""
+echo "=== Generating voice guidance (Edge TTS neural voices → mp3) ==="
+
+# Chinese — Xiaoxiao (warm, mature female)
+python3 -m edge_tts --voice "zh-CN-XiaoxiaoNeural" --rate="-40%" --text "继续专注，你做得很好。" --write-media "$AUDIO_DIR/mid-guidance-zh.mp3"
+echo "  Created: mid-guidance-zh.mp3"
+
+python3 -m edge_tts --voice "zh-CN-XiaoxiaoNeural" --rate="-40%" --text "练习完成。请记录你的感受。" --write-media "$AUDIO_DIR/end-guidance-zh.mp3"
+echo "  Created: end-guidance-zh.mp3"
+
+# English — Aria (neutral, professional female)
+python3 -m edge_tts --voice "en-US-AriaNeural" --rate="-40%" --text "Stay focused. You're doing well." --write-media "$AUDIO_DIR/mid-guidance-en.mp3"
+echo "  Created: mid-guidance-en.mp3"
+
+python3 -m edge_tts --voice "en-US-AriaNeural" --rate="-40%" --text "Practice complete. Take a moment to reflect." --write-media "$AUDIO_DIR/end-guidance-en.mp3"
+echo "  Created: end-guidance-en.mp3"
+
+echo ""
+echo "=== All audio assets generated in $AUDIO_DIR ==="
+ls -lh "$AUDIO_DIR"