ComfyUI Node: 🎭 Character Voices

🎭 Character Voices:

The CharacterVoicesNode is a specialized component within the TTS Audio Suite designed to manage voice references for text-to-speech (TTS) applications. Its primary purpose is to provide a unified approach to handling voice data, including both audio and text, across various TTS engines. This node facilitates the integration of character-specific voices by allowing users to specify a voice name and optionally provide reference text or audio input. It supports loading voice data from predefined folders or directly from user input, ensuring flexibility in how voice data is sourced and utilized. The node is particularly beneficial for applications that require consistent voice characterizations, such as storytelling or interactive dialogue systems, by ensuring compatibility with multiple TTS engines like F5-TTS and ChatterBox. By managing both audio and text references, the CharacterVoicesNode enhances the richness and authenticity of generated speech, making it a valuable tool for AI artists and developers working on voice-driven projects.

🎭 Character Voices Input Parameters:

voice_name

The voice_name parameter specifies the name of the voice to be used. It determines the source of the audio and text data, which can be loaded from a predefined folder if the name is not "none". This parameter is crucial as it directly influences the character of the voice output. There are no explicit minimum or maximum values, but it should correspond to a valid voice identifier within the system.

reference_text

The reference_text parameter allows you to provide a text reference that the node can use to generate speech. If not provided, the node attempts to use text from the voice folder. This parameter impacts the content of the speech output and is essential for applications requiring specific dialogue or narration. There are no explicit constraints on the text length, but it should be relevant to the intended speech output.

opt_audio_input

The opt_audio_input parameter is an optional input that allows you to provide a direct audio source. This can be used when you have a specific audio file that you want to use instead of relying on the predefined voice folders. It offers flexibility in sourcing audio data, especially when custom or unique audio inputs are required. The parameter should be a valid audio file path or object.

🎭 Character Voices Output Parameters:

narrator_voice_data

The narrator_voice_data output is a structured data object containing the audio tensor, audio path, reference text, character name, and source information. This output is crucial as it encapsulates all the necessary data for generating speech, ensuring that the TTS engine has access to both the audio waveform and the associated text. It provides a comprehensive package of voice data that can be used across different TTS engines.

character_name

The character_name output provides the name of the character associated with the voice data. This is important for applications that need to maintain consistency in character voices, such as in storytelling or interactive applications. It helps in identifying and managing different voice profiles within a project.

🎭 Character Voices Usage Tips:

• Ensure that the voice_name corresponds to a valid voice profile within your system to avoid errors and ensure the correct voice is loaded.
• When providing reference_text, make sure it is relevant and concise to improve the clarity and quality of the generated speech.
• Utilize the opt_audio_input parameter for custom audio needs, especially when working with unique or non-standard voice data.

🎭 Character Voices Common Errors and Solutions:

⚠️ Character Voices: Voice file not found: <voice_name>

• Explanation: This error occurs when the specified voice_name does not correspond to an existing audio file in the system.
• Solution: Verify that the voice_name is correct and that the corresponding audio file exists in the expected directory.

⚠️ Character Voices: No voice specified - provide voice_name or opt_audio_input

• Explanation: This error indicates that neither a voice_name nor an opt_audio_input was provided, leaving the node without a source for voice data.
• Solution: Ensure that you specify either a valid voice_name or provide an opt_audio_input to supply the necessary voice data.