ComfyUI-Impact-Pack: Update - Detailer with Differential Diffusion

The "noise_mask_feather" feature in the Detailer function of the Impact Pack has been improved.

"noise_mask_feather" applies feather to the noise_mask used in the i2i process.
The issue is that it wasn't very effective.

[neonelll]

You are doing great man.
I still wish you had a mic or could use text-to-speech AI <3

[ultimategolfarchives]

We don't hear much of this new feature and creating pngs, but this is major updates.

Great video ❤

[michail_]

I have a question for you. I tried to apply a Detailer (which splits an image into many squares) to AnimateDiff. And for some reason every square starts generating a character. Tried manipulating denoise and cfg but it doesn't work. But that's with the input video. And when I just generate text2video, everything is fine. Is it possible I'm wiring something wrong? Can you please tell me how to avoid this?
Thank you.

[drltdata]

You can download from ComfyUI from here:

And ComfyUI Impact Pack extension from here:

And workflow from here:

[NotThatOlivia]

how big the "noise_mask_feather" in comparison to "max_size" should be?

let's say that I have max_size = 1280 (for SDXL workflow), then noise_mask_feather = ???

[OutOfWards]

Does the Impact-Pack support nested wildcards?

[harshitpruthi]

hey, i am running comfyui on google colab, no matter how many times i update comfyui the differntial diffuor is not coming, what should i do

[eucharistenjoyer]

Great addition! Is this supposed to work on MaskDetailer? I updated mine and it doesn't seem to behave the same.

[jk]

why the hell is there no sound???? REALLY DUDE. Are your freaking KIDDING ME!!!