sunfounder_tts

Text-to-Speech brick — synthesizes speech from text and plays it through the system audio device.

Engines

Engine	Quality	Online	Notes
EdgeTTS (default)	Natural	Yes	Microsoft cloud TTS, free, 100+ voices across 50+ languages
Espeak	Robotic	No	Compact formant synthesizer, zero runtime dependencies
OpenAI_TTS	Best	Yes	OpenAI cloud TTS, 6 voices, needs OPENAI_API_KEY

Usage

EdgeTTS — recommended default (free cloud TTS)

from sunfounder_tts import EdgeTTS

tts = EdgeTTS(voice="zh-CN-XiaoxiaoNeural", gain=0.50)
tts.say("你好，我是UNO")

Available voices:

voices = EdgeTTS.available_voices()
# {'zh-CN-XiaoxiaoNeural': 'CN female, natural',
#  'zh-CN-YunxiNeural': 'CN male, professional',
#  'en-US-AriaNeural': 'US female, natural', ...}

tts.set_voice("en-US-JennyNeural")
tts.set_gain(0.75)   # volume factor (1.0 = original)

Espeak — compact offline

from sunfounder_tts import Espeak

tts = Espeak()
tts.set_voice("en-us+f3")   # American English female
tts.set_amp(200)            # volume 0-200, default 100
tts.set_speed(150)          # speed 80-260, default 175
tts.set_pitch(50)           # pitch 0-99, default 50
tts.say("Hello world")

OpenAI TTS — cloud

from sunfounder_tts import OpenAI_TTS

tts = OpenAI_TTS(api_key="sk-...")
tts.set_voice(tts.Voice.ALLOY)
tts.say("Hello from the cloud")

Audio Output

Playback via ALSA LineOut (hw:0,1) using PyAudio. The Qualcomm Codec mixer is configured at container startup by setup_audio_output() in robot_shield.

EdgeTTS outputs MP3; AudioPlayer auto-detects MP3 via magic bytes (0xFF 0xE0) and converts to WAV via sox before playback.

System-wide ALSA config written to /etc/asound.conf at image build time:

pcm.!default { type plug slave.pcm { type hw card 0 device 1 } }
ctl.!default { type hw card 0 device 1 }

Dependencies

• EdgeTTS: edge-tts pip package
• Espeak: espeak system package
• OpenAI_TTS: requests pip package + internet