The Ultimate ComfyUI Img2Img Workflow: SDXL All-in-One Guide! 💪

[@alwilson6471]

Perfection. I love your teaching/demo technique and, and others have commented, so generous to provide the workflow diagrams free whereas others only do so for patreon members.

[@illuminativr2286]

Finally, a workflow I can simply download for FREE!!! 🙌🙌🙌🙌 It works AWESOME btw, KEEP IT UP my man! 💯

[@RealityCheckVR]

🙌 Glad it helped!

[@webraptor007]

Thank you so much....excellent work. Keep it up. Looking forward for more!

[@MadHaTr4LiFe]

As an older computer user having your AI folders in the A: drive tripped me out. Great tutorial. Thank you. You can also drag and drop those images from browser tab to tab, so no need to save them anywhere when wanted to load up a workflow. (Tested in Edge)

[@RealityCheckVR]

Thanks for sharing! 🙌

[@darkkwod]

11:59 , now I get how it works, you are doing fine work, TY.

[@RealityCheckVR]

Sorry it's so long, I was tired lol. I'll have some updates this week when I have some free time from work! 🙌

[@darkkwod]

@@RealityCheckVR It's long because it's well explained, no need to say sorry at all. For me at least it was the first video that made me actually understand the img2img.

[@ferniclestix]

reee, its so tangly! cant stand cramped layouts but your great at explaining it all XD nice work, imma steal your sizing note for my next tutorial muahhaha!😋 keep up the good work :D

[@RealityCheckVR]

It's built in a big spaghetti web, I just condense it when I put it out for others so it's simple to use, just not easy to dissect unfortunately. Sharing is caring! 🙌

[@daan3898]

great node setup to learn more about comfy, thanks for the workflow ❤‍🔥

[@PowerHouseWealth]

Thank you for this my good sir 🙌🏻

[@SergeGolikov]

Recently got into Comfy and this was super useful, thank you🍷

[@RealityCheckVR]

Glad it was helpful!

[@embdigitality]

This is an amazing workflow, thank you!!!

[@RealityCheckVR]

Glad you like it! Working with any new ones lately?

[@embdigitality]

@@RealityCheckVR I'm exited to experiment, switching back and forth between ComfyUI and A1111, they are both amazing for different tasks!

[@jaredbeiswenger3766]

You have an A: drive? You are a madman.

[@bl1tz_x]

Thank you, Thor! 😊

[@RealityCheckVR]

Hurrah!! 🔨 Let's make some AI ART! 🤘

[@ysy69]

Excellent tutorial, keeping it simply and straight forward.

[@RealityCheckVR]

🙌

[@Hackmond]

Thank you!

[@RealityCheckVR]

You're Welcome! 😎😎

[@jjcs2917]

HI. Thanks! Derfuu custom model not available anymore in Civitai. Can't seem to find anywhere else too.

[@ThatGreenSpy]

Same.

[@MayorAwesome]

Great video!

[@RealityCheckVR]

Is it.. Is it really you?? It's been so long.. I've been so... so cold.. without.. you. 🤗 LoL How are you doing @MayorAwesome? missed you broski!

[@MayorAwesome]

@@RealityCheckVR Heheh. I've been lurking for a while. :) Good to see you're still at it.

[@3d-illusions]

When I drag the image from the imgur link you provided, it says 'unable to find workflow in edFdX79.png' , what's the solution to this please?

[@TheRedline3]

Hello, When I try to use the IMG2IMG workflow I get the following error. Do you know the solution? When loading the graph, the following node types were not found: Image scale to side Integer Nodes that have failed to load will show as red on the graph.

[@suyy_00]

Nice job man. Really appreciate your video. Keep up the good work. You earned my sub and gonna keep checking your channel. :)

[@RealityCheckVR]

Thanks brother)

[@HipPorRing]

how does the workflows automatically appear when dropping image into comfyui?

[@Queenbeez786]

the derfuu link doesnt work anymore

[@kernsanders3973]

The example workflow doesnt work when drag and dropping it into ComfyUI, Imgur strips PNGs of their metadeta when uploaded. Others upload their workflow PNGs to google drive, or the json file would be appreciated. Manually trying to rebuild someones workflow is a nightmare

[@RealityCheckVR]

thanks for the comment! I'll make some updates soon!

[@robinmountford5322]

Thanks !!!!

[@rtuesday2111]

Prompt outputs failed validation LoadImage: - Custom validation failed for node: image - Invalid image file: 00005-1840926806 (1).png UpscaleModelLoader: - Value not in list: model_name: '4x-UltraSharp.pth' not in [] LoraLoader: - Value not in list: lora_name: 'sd_xl_offset_example-lora_1.0.safetensors' not in [] any idea how to fix it ?

[@BrandinoTheFilipino]

the Derfuu math page does not exist where can i find it?

[@IrisSelene3]

I get two failed to load nodes: When loading the graph, the following node types were not found: Image scale to side Integer

[@RealityCheckVR]

Hey there, I've updated the google drive with the 4x sharpen.pth files which will help if you didn't have them for the upscaler. Did you download the Derfuu custom node?

[@IrisSelene3]

@@RealityCheckVR I already fixed it, the folder name was wrong

[@kalu_link]

I seem to have the same issue but I'm not able to solve it. DO you have any idea why? Great tutorial though! Thanks for sharing your knowledge!

[@GarethOakey]

I have the same issue, which folder name is wrong?

[@MrArelund]

@@GarethOakey If you downloaded it from Defruu's github, you need to remove '-master' from the end of the foldername to just 'Derfuu_ComfyUI_ModdedNodes' and restart ComfyUI

[@tranvan3999]

thank you

[@tortoise4529]

I could not achieve a good result. TXT2IMG gives me a pattern with robots on a purple background instead of the body of the robot, as shown in the video. I tried to use your Workflow in IMG2IMG mode with my Prompt, but I got the picture even more bad than I get in my Workflow without a SDXL model.

[@kennash497]

In this scenario, where do you put the lora in the pipeline?

[@_1_sebastian_1_]

well, I do not want to save the image first. I want to use it directly to refine it. In one step.

[@RealityCheckVR]

can you write out your perfect workflow that would be useful to you? Are you doing txt2image or image2image?

[@Trexatron]

Hey Reality Check VR, I saw this video and found it very helpful, so I sent some money your way as a thank you. I am stuck on one thing though, if you have any advice. I want to incorporate the 'Cutoff' nodes from BlenderNeko into this workflow, but I haven't been able to figure it out on my own. I don't know how to substitute the Cutoff nodes with the existing positive and negative prompts without messing up the VAE and Refiner decodes. Any way to do this without breaking the whole thing? Thanks.

[@RealityCheckVR]

Hey man! I'd be happy to help, I haven't had a chance to play with the cutoff nodes just yet, send me a workflow of what you have and I will see what I can discover! I've got a Patreon started and a paypal link in my description, I haven't seen anything come in and I certainly don't expect anything from anyone so no worries! 😎

[@avkna1830]

how do I put Derfuu Custom Nodes on google collab comfy ui

[@ilokikoval]

The custom nodes link is dead

[@crax83]

there's a github repo

[@RealityCheckVR]

can you link here?> I'll add it in!

[@Murderface666]

Derfuu is gone.

[@yiluwididreaming6732]

Bro I've got a 3050. 4gbvram (ASUS TUF gaming) laptop. This workflow, when using Turbo model is lightning quick {5 seconds). One of the reasons i use Comfy specifically for SDXL.

[@RealityCheckVR]

Any new workflows lately?