YuE2 RVC 翻唱:
YuE2RVCCover is a node designed to facilitate the transformation of a song's audio using a specified voice model, allowing for the creation of a cover version with a different vocal timbre. This node is particularly useful for artists and creators who wish to experiment with different vocal styles or create unique renditions of existing songs. By leveraging the capabilities of the YuE2 framework, YuE2RVCCover provides a seamless way to apply voice conversion techniques, ensuring that the resulting audio maintains high quality and artistic integrity. The node's primary goal is to enable users to explore creative possibilities in music production by altering the vocal characteristics of a song while preserving its original melody and accompaniment.
YuE2 RVC 翻唱 Input Parameters:
voice
This parameter specifies the voice model to be used for the cover. It determines the vocal timbre that will be applied to the song audio. The voice model is selected from available YuE2 RVC voices, which are pre-trained to provide various vocal characteristics.
song_audio
This parameter represents the original audio of the song that you wish to transform. It serves as the input audio that will undergo the voice conversion process. The quality and clarity of the input audio can significantly impact the final output.
semitone_shift
This integer parameter allows you to adjust the pitch of the song by shifting it up or down in semitones. The range is from -12 to 12, with a default value of 0, meaning no pitch shift. Adjusting this can help match the vocal range of the selected voice model or achieve a desired artistic effect.
index_rate
This float parameter, ranging from 0.0 to 1.0 with a default of 0.75, controls the rate at which the voice model's index is applied to the song. A higher value results in a stronger application of the voice model's characteristics, while a lower value retains more of the original vocal qualities.
protect
This float parameter, with a range from 0.0 to 0.5 and a default of 0.33, determines the degree of protection applied to the original vocal characteristics. A higher value means more of the original voice is preserved, which can be useful for maintaining certain vocal nuances.
vocal_gain_db
This float parameter allows you to adjust the gain of the vocal track in decibels, ranging from -18.0 to 12.0 with a default of 0.0. This can be used to balance the vocal volume relative to the accompaniment or achieve a specific mix.
accompaniment_gain_db
Similar to vocal_gain_db, this float parameter adjusts the gain of the accompaniment track in decibels, with the same range and default value. It helps in balancing the overall mix by controlling the volume of the non-vocal elements.
YuE2 RVC 翻唱 Output Parameters:
AUDIO
This output provides the transformed audio file, which is the result of applying the selected voice model to the input song audio. It contains the new vocal rendition while preserving the original melody and accompaniment.
YUE2_RESULT
This output contains metadata and additional information about the transformation process, such as the job ID and status. It is useful for tracking and managing the conversion tasks within the YuE2 framework.
STRING
This output provides a JSON string representation of the transformation status and results. It includes detailed information about the process, which can be useful for debugging or further analysis.
YuE2 RVC 翻唱 Usage Tips:
- Experiment with different semitone_shift values to find the best pitch match for the selected voice model, enhancing the naturalness of the cover.
- Use the index_rate parameter to fine-tune the balance between the original and transformed vocal characteristics, achieving the desired artistic effect.
- Adjust the vocal_gain_db and accompaniment_gain_db parameters to ensure a well-balanced mix, especially when integrating the cover into a larger musical project.
YuE2 RVC 翻唱 Common Errors and Solutions:
"RVC 音色已被移动或删除，请刷新 ComfyUI 页面"
- Explanation: This error occurs when the selected voice model is no longer available, possibly due to changes in the voice database.
- Solution: Refresh the ComfyUI page to update the list of available voice models and select a different one.
"该音色没有说话人 {sid}；可用说话人：{names or '无'}"
- Explanation: The specified speaker ID does not exist for the selected voice model, indicating a mismatch between the voice and speaker selection.
- Solution: Review the available speakers for the chosen voice model and select a valid speaker ID from the provided list.