A simple workflow to clone a voice from audio file.
Voices in this video were created with this workflow.
IMPORTANT: The audio file *** MUST BE LESS THAN 15 SECONDS **
Clone any voice. All you need it about 10 to 14 seconds of audio. Don't go above 14 seconds. If you do it puts some extraneous audio into your final output file.
Also make sure it is good quality audio.
This workflow does a pretty good job of rendering a voice that sounds like the real person.
Install the F5-TTS custom node.
Install the Transcription node
https://github.com/SWivid/F5-TTS
https://github.com/royceschultz/ComfyUI-TranscriptionTools/tree/master
Please look in "output" folder for the audio file (FLAC)
Workflow by trashkollector