ComfyUI Node: 📝 Phoneme Text Normalizer

📝 Phoneme Text Normalizer:

The PhonemeTextNormalizer is a versatile node designed for multilingual text-to-speech (TTS) preprocessing. Its primary function is to normalize text by handling special characters, phonemization, and various text transformations, making it an essential tool for preparing text for TTS applications. This node can automatically detect the language of the input text and apply appropriate normalization techniques, such as Unicode decomposition, IPA phonemization, and character mapping. By doing so, it ensures that the text is in a suitable format for further processing in TTS systems. The PhonemeTextNormalizer is particularly beneficial for artists and developers working with diverse languages, as it simplifies the text normalization process and enhances the quality of TTS outputs.

📝 Phoneme Text Normalizer Input Parameters:

text

The text parameter is the input string that you want to normalize. It is the raw text that will undergo various normalization processes depending on the selected method. This parameter is crucial as it directly influences the output of the node. There are no specific minimum or maximum values for this parameter, but it should be a valid string.

method

The method parameter determines the normalization technique to be applied to the input text. Options include "Pass-through," "Unicode Decomposition," "IPA Phonemization," and "Character Mapping." Each method offers a different approach to text normalization, affecting the final output. The default value is typically "Pass-through," which leaves the text unchanged.

language

The language parameter specifies the language of the input text. It can be set to "Auto-detect" to let the node automatically determine the language, or you can manually specify a language. This parameter is important for methods like IPA Phonemization, which require language-specific processing. There are no strict minimum or maximum values, but it should be a valid language code or "Auto-detect."

show_debug

The show_debug parameter is a boolean flag that, when set to true, enables the display of debug information. This includes details about the normalization process, such as the method used, the detected language, and any changes made to the text. The default value is typically false, meaning debug information is not shown unless explicitly requested.

📝 Phoneme Text Normalizer Output Parameters:

normalized_text

The normalized_text parameter is the output of the node, representing the text after normalization. It reflects the transformations applied based on the selected method and language settings. This output is crucial for ensuring that the text is in an optimal format for TTS systems, enhancing the clarity and accuracy of the synthesized speech.

📝 Phoneme Text Normalizer Usage Tips:

• Use the "Auto-detect" option for the language parameter when working with multilingual text to ensure the correct language-specific normalization is applied.
• Enable the show_debug parameter to gain insights into the normalization process, which can be helpful for troubleshooting and understanding how different methods affect the text.
• Experiment with different method options to find the most suitable normalization technique for your specific TTS application, as each method offers unique benefits.

📝 Phoneme Text Normalizer Common Errors and Solutions:

"⚠️ Both phonemizers failed - standard: {e1}, windows: {e2}"

• Explanation: This error occurs when both the standard and Windows phonemizers fail to process the text.
• Solution: The node automatically falls back to Unicode decomposition in this case. Ensure that the input text is valid and consider using a different normalization method if the issue persists.

"Text length changed: {len(original_text)} → {len(normalized_text)} characters"

• Explanation: This message indicates that the length of the text has changed during normalization, which may affect alignment in TTS systems.
• Solution: Review the changes made to the text and adjust the input or method as necessary to maintain the desired text length.