Chinese TTS Voice Generation

Generate natural Chinese speech using Microsoft Edge TTS and send as Feishu voice messages.

Quick Reference

# 1. Generate MP3
LANG=en_US.UTF-8 LC_ALL=en_US.UTF-8 /home/clawpi/.local/bin/edge-tts \
  --voice zh-CN-YunxiNeural \
  --text "你的文本内容" \
  --write-media /tmp/output.mp3

# 2. Convert to Opus
ffmpeg -i /tmp/output.mp3 -c:a libopus -b:a 64k -ar 48000 \
  /home/node/.openclaw/workspace/output.opus -y

# 3. Send via Feishu
message(asVoice=true, contentType="audio/ogg",
        filePath="/home/node/.openclaw/workspace/output.opus",
        filename="output.opus")

Critical Rules

1. Always set UTF-8 env vars — System locale is ISO-8859-1, Chinese text will be corrupted without LANG=en_US.UTF-8 LC_ALL=en_US.UTF-8
2. Must use opus format — Feishu only treats .opus as voice messages; MP3/WAV become file attachments
3. Must send from workspace — /tmp is not in Feishu's mediaLocalRoots whitelist; files there fail silently and fall back to plain text
4. Use ASCII filenames — Chinese filenames may cause encoding issues in multipart uploads

Voice Options

For Long Text

Write text to a file and use -f flag:

echo "长文本内容..." > /tmp/text.txt
LANG=en_US.UTF-8 LC_ALL=en_US.UTF-8 /home/clawpi/.local/bin/edge-tts \
  --voice zh-CN-YunxiNeural -f /tmp/text.txt --write-media /tmp/output.mp3

Troubleshooting

• Gibberish audio → Missing UTF-8 env vars
• File sent as attachment, not voice → Not opus format, or not from workspace path
• Upload fails silently → File not in mediaLocalRoots (use workspace dir)
• "哈米" instead of "虾米" → Google TTS (gtts) issue; use edge-tts instead