SipPulse AI - Text to Speech Conversion

Text to Speech Interface

The SipPulse AI platform provides powerful capabilities for converting text into natural-sounding speech. This section documents how to use the Text to Speech tool to transform written content into audio.

Accessing the Text to Speech Interface

1. Navigating to the Tool:

   • Access the left sidebar menu
   • Click on "Playground" to expand the options
   • Select "Text to Speech" to open the conversion interface

2. Main Interface Components:

   • Text Input Area: Central zone for entering the text you want to convert
   • Audio Visualization: Waveform display of the generated audio
   • Playback Controls: Play button, download option, and timeline
   • Generated Audio List: Right panel showing previously generated audio files

Configuration Options

1. Model and Voice Settings:

   • Model: Select from available TTS models (e.g., "eleven-labs-flash")
   • Language Options: Choose "Multi-language" or select a specific language
   • Voice Selection: Select from different voice options (e.g., "Aria")
   • Format: Choose the output audio format (e.g., "mp3")

2. Advanced Features:

   • Autoplay: Option to automatically play audio after generation
   • Voice Customization: Some models allow adjusting voice characteristics
   • Experimental Features: Access to new capabilities (tagged as "Experimental")

3. Keyboard Shortcuts:

   • Press "Ctrl+Enter" to quickly generate speech from the entered text
   • Use playback controls to manage audio playback

Speech Generation Process

1. Text Preparation:

   • Enter the text you want to convert into the text input area
   • Format your text with appropriate punctuation for natural speech patterns
   • Consider using SSML tags for advanced voice control (if supported by the selected model)

2. Configuration Selection:

   • Choose the appropriate model for your needs
   • Select the desired voice that matches your content
   • Configure language settings if generating speech in specific languages

3. Generating Speech:

   • Click the "Speak" button or use the Ctrl+Enter shortcut
   • The system processes your text and generates the audio
   • The waveform visualization displays once processing is complete

4. Review and Export:

   • Listen to the generated audio using the playback controls
   • Download the audio file using the download button
   • Regenerate with different settings if needed

API Integration

SipPulse AI provides a RESTful API for integrating text-to-speech capabilities directly into your applications. Below are examples of how to use the API in different programming languages.

API Parameters

• model: Specifies the TTS model (e.g., eleven-labs-flash)
• voice: Determines which voice to use (e.g., aria)
• format: Output audio format (e.g., mp3)
• api-key: Your SipPulse API authentication key

Python Example

import requests

url = 'https://api.sippulse.ai/tts/synthesize'
headers = {
    'accept': 'application/json',
    'content-type': 'application/json',
    'api-key': '$SIPPULSE_API_KEY'
}
data = {
    'text': "Let's see if it works",
    'model': 'eleven-labs-flash',
    'voice': 'aria',
    'format': 'mp3'
}

response = requests.post(url, json=data, headers=headers)

if response.status_code == 200:
    with open('output.mp3', 'wb') as f:
        f.write(response.content)
    print("Audio file created successfully!")
else:
    print(f"Error: {response.status_code}")
    print(response.text)

Node.js Example

const fetch = require('node-fetch');
const fs = require('fs');

const url = 'https://api.sippulse.ai/tts/synthesize';
const data = {
  text: "Let's see if it works",
  model: 'eleven-labs-flash',
  voice: 'aria',
  format: 'mp3'
};

const options = {
  method: 'POST',
  headers: {
    'accept': 'application/json',
    'content-type': 'application/json',
    'api-key': '$SIPPULSE_API_KEY'
  },
  body: JSON.stringify(data)
};

fetch(url, options)
  .then(response => {
    if (!response.ok) {
      throw new Error(`HTTP error! Status: ${response.status}`);
    }
    return response.buffer();
  })
  .then(buffer => {
    fs.writeFileSync('output.mp3', buffer);
    console.log('Audio file created successfully!');
  })
  .catch(error => console.error('Error:', error));

cURL Example

curl -X 'POST' \
  'https://api.sippulse.ai/tts/synthesize' \
  -H 'accept: application/json' \
  -H 'content-type: application/json' \
  -H 'api-key: $SIPPULSE_API_KEY' \
  -d '{
    "text": "Let'\''s see if it works",
    "model": "eleven-labs-flash",
    "voice": "aria",
    "format": "mp3"
  }' \
  --output output.mp3

Usage Considerations

1. Voice Quality Optimization:

   • Keep sentences at a natural length for more realistic speech patterns
   • Use appropriate punctuation to control pacing and intonation
   • Test different voices to find the best match for your content

2. Cost Management:

   • Be aware that costs typically scale with the length of the text
   • Consider breaking very long texts into smaller segments
   • Use the appropriate model based on quality vs. cost requirements

3. Performance Factors:

   • Higher quality models may have longer processing times
   • Very long texts will take more time to process
   • Some voices may be optimized for specific languages or content types

4. Content Guidelines:

   • Avoid generating speech with sensitive personal information
   • Be aware of usage policies regarding impersonation or deception
   • Follow local regulations regarding synthetic voice generation

By using the SipPulse AI Text to Speech tool, you can efficiently convert written content into natural-sounding speech for a wide range of applications, including voice assistants, educational content, accessibility features, and more.