Comfy_IrodoriTTS is a custom node designed for use with ComfyUI that integrates the Irodori-TTS text-to-speech model and its fork, Emoji-TTS. This tool enables users to generate speech from text while allowing for various configurations, including speaker reference from audio files, VoiceDesign captions, and adjustments for CFG, Rescale, and LoRA applications.
- Supports loading IrodoriTTS checkpoints and generating speech from text.
- Allows for speaker specification using reference audio files and includes a user-friendly emoji picker.
- Provides advanced configuration options for CFG, duration, and audio trimming to enhance the quality of generated speech.
Context
The Comfy_IrodoriTTS tool serves as a bridge between ComfyUI and the Irodori-TTS model, facilitating the generation of high-quality speech from textual input. Its purpose is to enhance the text-to-speech capabilities within the ComfyUI environment, enabling users to leverage advanced features for customized audio outputs.
Key Features & Benefits
This custom node includes several practical features that significantly improve the user experience:
- Checkpoint Loading: Users can easily load various IrodoriTTS checkpoints to access different voice models.
- Reference Audio Integration: The ability to specify reference audio allows users to create speech outputs that closely match a desired speaker's voice or tone.
- VoiceDesign Support: Users can define voice characteristics and speaking styles through captions, providing greater control over the generated audio.
Advanced Functionalities
The tool includes specialized capabilities such as:
- LoRA Integration: Users can apply Low-Rank Adaptation (LoRA) models to enhance the speech output further, allowing for adjustments in voice characteristics.
- Detailed CFG Settings: Advanced users can fine-tune the conditional guidance factors (CFG) to control the influence of text, speaker, and voice characteristics on the generated audio.
- Automatic Duration Estimation: For the v3 model, automatic estimation of the speech duration based on the input text is available, simplifying the process for users.
Practical Benefits
Comfy_IrodoriTTS streamlines the workflow for generating text-to-speech outputs in ComfyUI by providing a user-friendly interface and versatile configuration options. This tool enhances control over audio quality and speaker characteristics, making it easier to produce high-quality, customized speech outputs efficiently.
Credits/Acknowledgments
The development of this custom node is based on the original Irodori-TTS and Emoji-TTS projects, with contributions from various authors. Users are encouraged to check the respective licenses for Irodori-TTS, Emoji-TTS, and the models utilized within this tool, as detailed in the repository's LICENSE file.
