Best multi-model voice generator - by far!

[DarkSoundLab]

Great job incorporating so many different models into one package. Yours is the only one I could actually get VibeVoice to work with. Thank you!

I am wondering if you could add or maybe explain how one might go about adding a way to change or adjust the pitch and speed of the generated output, relative to the input voice or voice sample. Is that possible to integrate into your workflow?
Additionally, I am having an issue with your other workflows not loading. I can only get Chatterbox and VibeVoice Integrations to load.

Also, I think you mentioned somewhere that you might be doing a full walkthough video of your project explaining usage, settings, etc.. Where would one find that posted?

[GeekyGhost]

Great job incorporating so many different models into one package. Yours is the only one I could actually get VibeVoice to work with. Thank you!

I am wondering if you could add or maybe explain how one might go about adding a way to change or adjust the pitch and speed of the generated output, relative to the input voice or voice sample. Is that possible to integrate into your workflow? Additionally, I am having an issue with your other workflows not loading. I can only get Chatterbox and VibeVoice Integrations to load.

Also, I think you mentioned somewhere that you might be doing a full walkthough video of your project explaining usage, settings, etc.. Where would one find that posted?

So, there is the TTS Text node and the TTS SRT node, you can actually control the speed fairly well with the SRT node. It lets you control the timings. You're essentially saying start saying this here, and then finish by here. Though this could also be achieved with traditional audio editing techniques.

[GeekyGhost]

A little, but not enough to diagnose. Which workflow specifically was giving the issue?

[DarkSoundLab]

As I said in the OP, I can only get Chatterbox and VibeVoice Integrations to load. I was mainly interested in the VoiceChanger but may whatever the issue is, is the same with the others.

[GeekyGhost]

Ahh, Okay, yeah, I was assuming it was the voice changer. That one I got working by switching to the 3.49 portable with 3.12. I was going to fix the 3.13 repo conflict for my install, but I had some other tools I needed 3.12 for so I just did a clean install and swapped over my stuff. I didn't dive in to deep, but it doesn't like 3.13 at all, and I'm unsure about 3.10 or 3.11, but I know it works well on 3.12, as long as you don't have reactor, but that can be fixed.

[DarkSoundLab]

So it can only run on the portable version with 3.12?
Is there any fix for using it with what I am on now?
Is it something in the workflow, the nodes or models, etc. that is causing them to not load? Like I said, I don't know too much about all this, but I'm trying.

[diodiogod]

@DarkSoundLab this should work great on a new portable python 3.13. You will only be missing some small features, like mediapipe for silent speech detection, and some Audio Separation models might not work, but the best ones do.

I recommend you to download a portable comfyui, install comyui manager. restart. then inside comfyui, search for this node and install it there. Then restart. It should run the install script and hopefully everything should just work (that is the idea anyway). If it doesn't you could open a bug issue on the issue tab and post your console error.

[GeekyGhost]

And keep in mind, post production doesn't have to be outside of comfyUI

[diodiogod]

Great job incorporating so many different models into one package. Yours is the only one I could actually get VibeVoice to work with. Thank you!

I am wondering if you could add or maybe explain how one might go about adding a way to change or adjust the pitch and speed of the generated output, relative to the input voice or voice sample. Is that possible to integrate into your workflow? Additionally, I am having an issue with your other workflows not loading. I can only get Chatterbox and VibeVoice Integrations to load.

Also, I think you mentioned somewhere that you might be doing a full walkthough video of your project explaining usage, settings, etc.. Where would one find that posted?

I did not have time to do the video yet, unfortunately.
About adjusting the pitch and speed, this would need to be a post generation thing, like I do with the TTS Srt node like @GeekyGhost suggested. I know that you can adjust pitch with RVC during a conversion, I wonder if you could use it without a voice conversion. But like I @GeekyGhost suggested just now you can use a lot of other audio manipulation nodes out there!

[GeekyGhost]

Yeah, Pitch shifts are pretty easy, just gotta make sure you don't speed it up or slow it down in the process lol.

s1audio (1).wav

ComfyUI_00056_ (1).mp3

That's a pitch shift of -5.1 just using regular techniques inside of comfyUI. You can bring a lot of traditional tools in. I'm even using chroma and luma removals to go with my AI background removal models!