Import Failed

[lunatico67]

Hi
I am on Windows 11 ; ComfyUI portable version.
Error message occurred while importing the 'ComfyUI F5-TTS' module.

Traceback (most recent call last):
File "F:\ComfyUI__LUNA\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\nodes.py", line 2113, in load_custom_node
module_spec.loader.exec_module(module)
File "", line 999, in exec_module
File "", line 488, in call_with_frames_removed
File "F:\ComfyUI__Moderno\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-F5-TTS_init.py", line 2, in
from .F5TTS import F5TTSAudio, F5TTSAudioInputs
File "F:\ComfyUI__LUNA\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-F5-TTS\F5TTS.py", line 22, in
from f5_tts.model import DiT,UNetT # noqa E402
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ModuleNotFoundError: No module named 'f5_tts'

[niknah]

A lot of people seem to be having problems with this.

I have updated this custom node so it tries to get the submodules two different ways. pygit2 + command line git.

The other way is to get the submodule manually by pasting this into the command line...

cd custom_nodes/ComfyUI-F5-TTS
git submodule update --init --recursive
pip install -r requirements.txt

[lunatico67]

Hi niknah, thanks for your information. I have updated your custom node and I don't get the import failed error anymore.
Please, ComfyUI-F5-TTS says ComfyUI node to make text to speech audio with your own voice.
I suppose it can also be done with other voices, not just yours, which is a voice clone.
Is there a video tutorial on how to do it?

[niknah]

Yes, get the voice audio, save it as a NameOfPerson.wav, write down what the person said and save it as NameOfPerson.txt
Put them both into the "input" folder in ComfyUI. Press "r" in ComfyUI and it should appear in the F5-TTS node.

Someone made a video here... https://www.reddit.com/r/StableDiffusion/comments/1id8spa/effortlessly_clone_your_own_voice_by_using/

[lunatico67]

Hi niknah
A lot of thanks for sharing the link of the video.
In the video he records his voice with the node Audio Recorder @ vrch.ai .
Is it possible, instead of recording your voice, to load any audio sample (I don't know what its length limit would be) in some node that loads audio samples, (load audio)?

[niknah]

Yes. Where is your audio from? Put the audio in a .wav file. In that workflow, replace the record audio node with a load audio node. Put the .wav file into the "input" folder in ComfyUI.

There are sample files in this repository in the examples/ folder. I've never used my own voice, just played around with other people's.

Whatever audio file you put in, F5-TTS will cut it off at about 15secs.