ComfyUI Node: Audio Cleaner

Audio Cleaner:

The AudioCleanupNode is a powerful tool designed to enhance the quality of audio generated by text-to-speech (TTS) systems. Its primary function is to improve the clarity and intelligibility of audio by removing unwanted noise and reverberation, which can often degrade the listening experience. By applying advanced audio processing techniques, this node effectively reduces background noise and echoes, allowing the main audio content, such as speech, to stand out more prominently. This enhancement is particularly beneficial for applications where clear and crisp audio is essential, such as in virtual assistants, audiobooks, and other TTS applications. The node leverages high-pass and low-pass filtering, noise reduction, and dereverberation techniques to achieve these improvements, making it an essential component for anyone looking to produce high-quality audio content.

Audio Cleaner Input Parameters:

audio

This parameter represents the audio input that the node will process. It should be in the ComfyUI audio format, which includes both the waveform and the sample rate. The audio input is the foundation upon which all enhancements are applied, and it is crucial for the node's operation.

denoise_strength

This parameter controls the intensity of the noise reduction process. It ranges from 0.1 to 1.0, with a default value of 0.5. A higher value results in more aggressive noise reduction, which can be useful in environments with significant background noise. However, setting it too high might affect the quality of the desired audio content.

dereverb_strength

This parameter determines the strength of the dereverberation process, which aims to reduce echoes and reverberation in the audio. It ranges from 0.0 to 1.0, with a default value of 0.7. Increasing this value can help in environments with a lot of echo, but excessive dereverberation might make the audio sound unnatural.

high_pass_freq

This parameter sets the cutoff frequency for the high-pass filter, which removes low-frequency noise from the audio. It ranges from 20.0 Hz to 500.0 Hz, with a default value of 100.0 Hz. Adjusting this parameter can help eliminate rumble and other low-frequency disturbances.

low_pass_freq

This parameter defines the cutoff frequency for the low-pass filter, which removes high-frequency noise. It ranges from 1000.0 Hz to 16000.0 Hz, with a default value of 8000.0 Hz. This setting is useful for reducing hiss and other high-frequency noise.

normalize

This parameter indicates whether the audio should be normalized, with options "true" or "false" and a default of "true". Normalization adjusts the audio to a consistent level, which can enhance the overall listening experience by ensuring that the audio is neither too quiet nor too loud.

Audio Cleaner Output Parameters:

enhanced_audio

This output parameter provides the processed audio in the ComfyUI audio format. The enhanced audio has undergone noise reduction, dereverberation, and filtering processes to improve its quality. This output is crucial for applications requiring high-quality audio, as it ensures that the audio is clear and free from unwanted noise and reverberation.

Audio Cleaner Usage Tips:

• To achieve the best results, start with the default settings and adjust the denoise_strength and dereverb_strength parameters incrementally based on the specific noise and reverberation levels in your audio.
• Use the high_pass_freq and low_pass_freq settings to target specific frequency ranges where noise is most prevalent, ensuring that the main audio content remains unaffected.
• Always listen to the processed audio to ensure that the enhancements have not introduced any artifacts or negatively impacted the desired audio content.

Audio Cleaner Common Errors and Solutions:

"输入音频格式不支持，应为ComfyUI的AUDIO类型"

• Explanation: This error occurs when the input audio is not in the expected ComfyUI audio format, which includes both waveform and sample rate.
• Solution: Ensure that the audio input is correctly formatted as a dictionary with "waveform" and "sample_rate" keys. Use the LoadAudio node to properly format your audio before processing.

"处理音频失败: <error_message>"

• Explanation: This error indicates a failure in the audio processing pipeline, possibly due to incorrect parameter values or unexpected audio content.
• Solution: Check the input parameters for any incorrect values and ensure that the audio content is suitable for processing. If the issue persists, review the error message for specific details and adjust the parameters accordingly.