ComfyUI Impact Pack - Tutorial #7: Advanced IMG2IMG using Regional Sampler

Today, I will introduce how to perform img2img using the Regional Sampler.

You can download from ComfyUI from here:

And ComfyUI Impact Pack extension from here:

Is there a way to use Regional prompting (or something similar) with Controlnet? Could help me specifying the eyes, or limbs when i load in the sketch of a character that i've drawn. Trying to use ComfyUI to color my sketches basically.

[alecubudulecu]

that's amazing...thanks for this quick tutorial. really appreciate all the work you do.

[AhmadAbukwas]

Thanks for great info, did you make mask colors in comfyui?

[xueqinghuang]

very good tutor and nice way to compose use your method! Thx for your Selfless sharing

[drltdata]

You can download from ComfyUI from here:

And extension from here:

Workflow: (you can drag drop this image)

[___x__x_r___xa__x_____f______]

is there a workflow for regional lora's that uses color masks to send one lora to each character?

[holchan]

How do you see the progress as the sampler is working?

[jedrzejtazbir]

When I try to reproduce this workflow, I get the following error: "Error occurred when executing RegionalPromptColorMask //Inspire:
not enough values to unpack (expected 4, got 3)". Do you have any idea as to what the issue might be? I created a color map following your other video.

[indiffident]

can't see the option to switch to dev, i have only: default, recent, legacy, forked. no dev. any help?

What does the image data in the Image Receiver do? I don't understand it.

[jabelardo]

Is is possible to use control net (lineart for example) combined with this method to keep further the original composition? or is this idea totally an overkill?