Feature Request: Configurable Speech Rate for TTS Providers

[jogi-poikola]

Not sure if feature ideas should be in discussions rather that in issues, but here it goes this time.

Problem

The speaking rate for Google Cloud TTS is currently hardcoded to 1.0 (normal speed) in the voice server. Users cannot adjust playback speed without modifying the source code.

Proposed Solution

Add configurable SPEAKING_RATE environment variable to control TTS playback speed:

# In $PAI_DIR/.env
SPEAKING_RATE=1.7  # Range: 0.25 to 4.0, default 1.0

Implementation:

// server.ts
const SPEAKING_RATE = parseFloat(process.env.SPEAKING_RATE || "1.0");

// In generateSpeechGoogle()
audioConfig: {
  speakingRate: SPEAKING_RATE,
  pitch: 0,
}

Benefits

• Users can adjust speech speed to personal preference
• Faster playback (1.5x - 2.0x) reduces notification time
• Slower playback (0.5x - 0.75x) improves comprehension for complex messages
• No code changes needed to adjust speed

Provider Support

• Google Cloud TTS: ✅ Supports 0.25x to 4.0x
• ElevenLabs: ❌ No rate control via API (uses default speed)

Default Value

Recommend keeping default at 1.0 (normal speed) for backward compatibility.

[orangemaster-pai]

Point your PAI at this and have them rebuild the server to use it and you can do more than just speed, you can control throughout the whole reply. Have your pai use the personality traits to tune. https://docs.cloud.google.com/text-to-speech/docs/ssml

[jogi-poikola]

I did this already to me personally, but I suggested that this sort of configuartion option could be included in the official PAI repo also. I am not sure if issues is really a right place to report this kind of ideas.