Transcribe Lyrics:
The PlenioTranscribeLyrics node is designed to transcribe sung words from audio recordings, specifically focusing on sections where singing is detected. Utilizing the faster-whisper large-v3 model, this node efficiently identifies and transcribes lyrics, aligning them with the beat grid of the final score. This functionality is particularly beneficial for music producers and AI artists who need to verify if the expected lyrics were sung correctly, section by section. By integrating this node into your workflow, you can ensure that the lyrical content of your recordings is accurately captured and organized, facilitating a seamless music production process.
Transcribe Lyrics Input Parameters:
audio
The audio parameter represents the audio recording from which the lyrics will be transcribed. It is crucial for the node's operation as it serves as the primary source of data for transcription. The quality and clarity of the audio can significantly impact the accuracy of the transcription results. Ensure that the audio is clear and free from excessive background noise for optimal performance.
expected_lyrics
The expected_lyrics parameter allows you to input the lyrics that you anticipate being sung in the recording. This parameter is used to verify the accuracy of the transcription by comparing the detected lyrics against the expected ones. If provided, the node will check each section to determine if the lyrics were sung as expected, offering insights into any discrepancies.
language
The language parameter specifies the language of the lyrics to be transcribed. This setting is essential for configuring the transcription engine to accurately process the audio. Selecting the correct language ensures that the nuances and phonetics of the lyrics are correctly interpreted, leading to more precise transcription results.
timeline
The timeline parameter is used to provide a reference timeline for the audio recording. If the timeline is associated with the same audio source, it helps in aligning the transcribed lyrics with the beat grid of the final score. This parameter is optional but can enhance the accuracy of the transcription alignment if available.
Transcribe Lyrics Output Parameters:
transcript
The transcript output provides the complete transcription of the sung lyrics detected in the audio recording. This output is crucial for reviewing and verifying the lyrical content of the recording, ensuring that the transcription aligns with the intended lyrics.
detected
The detected output indicates the sections of the audio where singing was found and transcribed. This information is valuable for identifying the specific parts of the recording that contain vocal content, allowing for targeted analysis and editing.
report
The report output offers a detailed summary of the transcription process, including any warnings or discrepancies found during the comparison with expected lyrics. This output is essential for understanding the accuracy and reliability of the transcription, providing insights into any sections that may require further attention.
Transcribe Lyrics Usage Tips:
- Ensure that your audio recordings are of high quality and free from excessive background noise to improve transcription accuracy.
- When using the expected_lyrics parameter, provide the most accurate and complete lyrics possible to facilitate effective verification and error detection.
- Select the correct language setting to ensure that the transcription engine accurately interprets the phonetics and nuances of the lyrics.
Transcribe Lyrics Common Errors and Solutions:
The transcription found no vocal melody in the source, so there are no sung words to transcribe.
- Explanation: This error occurs when the node does not detect any vocal melody in the audio recording, which may indicate that the source is instrumental or lacks clear vocal content.
- Solution: Verify that the audio recording contains vocal content. If the source is indeed instrumental, consider selecting the 'instrumental' option in the Cover Brief or manually entering the lyrics in the Song Sheet.
The timeline belongs to another recording; it was not used.
- Explanation: This warning indicates that the provided timeline does not match the audio source, preventing its use in aligning the transcription.
- Solution: Ensure that the timeline corresponds to the same audio recording being transcribed. If necessary, update the timeline to match the correct audio source.