Add Points and Synthesize is a Python-based tool designed to help language learners improve their pronunciation and comprehension by listening to texts in various voices. It ensures natural speech flow by automatically formatting text to match the optimal input format for text-to-speech (TTS) engines. The project utilizes Microsoft Speech Services to generate high-quality audio with diverse voices across multiple languages.
When learning a new language, it can be challenging to hear how a text should be read naturally. Many text-to-speech (TTS) services, including Google Translate, often produce monotonous, robotic voices, making listening practice less engaging.
Through exploration, we discovered a TTS service that required each sentence to end with a period and a space to ensure proper pausing. To automate this formatting, a spreadsheet was initially created, but copying large amounts of text became a challenge due to mobile security restrictions. This led to the development of this Python tool, which:
- Formats text by adding necessary punctuation and spacing.
- Uses Microsoft Speech Services to generate natural speech.
- Provides randomized voices for a more dynamic listening experience.
- Automatic Text Formatting: Ensures each sentence ends with a period and a space.
- Multi-language Support: Works with multiple languages including English, Portuguese, Spanish, French, German, and more.
- Microsoft Speech Integration: Fetches high-quality voices from Microsoft's AI-powered speech synthesis service.
- Randomized Voice Selection: Makes listening more engaging by selecting a different voice each time.
- MPV Integration for Audio Playback: Plays the generated speech automatically on supported platforms.
The tool currently supports the following languages (with at least 4 voice variations available):
- Arabic
- Chinese
- Dutch
- English
- French
- German
- Indonesian
- Italian
- Japanese
- Korean
- Polish
- Portuguese
- Russian
- Spanish
- Thai
- Turkish
- Vietnamese
Before using this tool, ensure you have the following installed:
- Python 3.x
- Termux (for Android users) (Optional, if running on a mobile device)
- MPV media player (Required for audio playback)
-
Clone the repository:
git clone https://github.com/sght500/add-points-synthesize.git cd add-points-synthesize -
Install dependencies:
pip install -r requirements.txt
-
Set up Microsoft Speech API credentials:
- Sign up for Microsoft Azure Speech Services.
- Retrieve your Subscription Key and Region.
- Export them as environment variables:
export SPEECH_KEY="your_subscription_key" export SPEECH_REGION="your_region"
-
Run the script:
python add-points-tts.py
- Enter your text line by line. Each sentence should be on a separate line.
- End input by entering an integer. This number specifies how many lines to delete from the end before processing.
- The tool automatically adds punctuation. It ensures that each sentence ends with a period and a space.
- Audio is generated using a random voice. The tool selects a voice from Microsoft Speech Services.
- MPV plays the generated audio file.
- Enter
0on the first line → Change language. - Enter
9on the first line → Exit the program.
Modify add-points.json to customize settings:
mpv_path: Path to MPV media player.target_languages: List of supported languages.min_voice_count: Minimum number of voices required for a language to be selectable.instructions: Language-specific prompts displayed in the program.
- GUI version for easier text input and playback.
- Offline TTS support to reduce reliance on Microsoft servers.
- Customization options for voice selection and speed adjustments.
Contributions are welcome! Feel free to submit issues, feature requests, or pull requests to improve the project.
This project is licensed under the AGPL-3.0 License. See LICENSE for details.
Developed with ❤️ to make language learning more engaging!