Stable Diffusion ComfyUI With Krita Installation Guide - Image Editor AI Image Integrated

hello everyone so today we are going to talk about how to use stable diffusions in photo editor this is a very cool open- Source photo editor called CR so you can go to their official website and download this installer and it will automatically detect your windows Mac OS or Linux and after install that you can click this icon open it and it is like that so in this tutorials we are going to show how to integrate stable diffusions using comfy UI with CR photo editors so you can do image Generations on cre and you can actually edit the photos in this photo editor eventually they also support realtime alike photo drawing using LCM model and which I will be covered that in another video but first we got to download this cre AI diffusions plugin from GitHub page and we are going to install it step by step using their installation method here I have comfy UI already and I am going to follow their instruction here the good thing is that it's only required 6 GB vram so first step copy the link and this is the folders directories that I unzip these files and then locating in this folder so basically I unzip it or drag and drop these two files into this G puck in folders and then you come back to CR image editor go to settings and you see the resource and you navigate down the python plug-in manager you will see the AI image diffusions plugin is already installed into here so you click this check box enable it press okay and then restart your Crea a photo image editor you will see the settings when you come back to configuration you check one more time make sure it's all set then you go to the Dockers you will see the AI image Generations checkbox right there that means you have already successful install this so we create a new empty image now we can enable this AI image Generations so when you enable this AI image generation you will see it on right side panel you will see this configurations button you can connect your com UI or if you don't have comfy UI you can click the local server managed by C install everything for you but since I have comy UI already then I will just use my existing comy UI right here when you see some custom nodes are missing you are required to install that in your comy UI by doing this start your comfy UI and then go to my comy UI manager I have to search this tooling custom node that I have to install in order to connect with cre image editor then we will click the install button in the config UI manager it's that easy simple just click the restart button and wait for the comfy UI to restart and once it's installed successfully we can minimize this web UI and command prompt windows and we just need to copy the IP address and the port numbers to CR as you can see there are still some errors and you can fix it one by one like in my case here I need to locate the IP adapter models files into the folder path that they want me to locate it because I know that I have the IP adapter models in my comyu but my folder path for SD 1.5 and sdxl is different so they have not recognized the model file so I'm going to come back to my comfyi folder and locate the directory as the CR show so in here the comy UI adapter plus I'm using and there's models files and back to create a click connect and then you see this connected successfully so after connected you see the uh AI image generation tab is right here I like to put the tab on top section you can start typing some simple prompts and like a beautiful castle and let's click generate and see what happened this is a very basic setup for this integration its alternative of using stable diffusion and comfi UI with the image editor you can further enhance some detail of an image now we don't need to use comy UI web browsers we just need to open up the command prompts here to let the backend comi running and as you can see it does load the models and the sampling steps here so right here we have the castle photos created for us and we can lower down the strength in this editor settings and there are some settings you can click over you can see stable diffusion features like control net open pose upscaler you can play around with and later I will try this control net open pose let's generate one more time I'm using the 80% strength and it creates something different and you can check the settings in here it's called the style present settings you can add Laura or choose a specific checkpoint models that you want to work with and in my example I'm going to use realistic Visions to create image and you can also add some Laura models in here Alo like here I have some detailer styler and I'm going to use ad you uh detail Laura model and I'm setting the strength as 60% and looks everything is okay when we are finished our settings here okay everything looks good it detects my display card here so yeah everything is fine then I will click okay to change the settings and then let's check again beautiful castle in landscape and then right in this drop- down menu you can click where the generate image are locating in your layer your photos file layer and then I click generate and wait for the result here so scroll down a little bit you can see the new result already here and there's a few result that I just clicked previously and uh well it's not very high details but then we can based on this image to enhance and doing more upscaling and further detailer enhancement with that so yeah this is how we can use stable diffusions comfy UI with CR image editor and here's few more Castle image that I generate just want to play around with and as you can see this looks pretty cool now we are going to play around with the control net open open pose so we can generate some character pose specified by us that we can using the control net open pose right here we can click this button and then we have chosen the pose icon and select the image layer well I should be creating a new layer for this uh the vector layer so let's create a vector layer on the layer sections and then we can choosing the vector layer to add open pose Skeleton on the screen so on the screen here we can drag and drop the open pose skeleton and we can move around the joints to make some pose that we want to do like this one I'm making like a character sitting down on something and then I can move the hands up a little bit just like waving say hi to people all right let's click generate and see okay so there's some new image um it looks a little bit dark but yeah the character have followed the open pose about 80% here well I have to define the backgrounds and the environment hopefully looks better a little bit so let's try another generation with defining the light on the backgrounds and let's see okay obviously they have a light on the background but then I want the character in places that is going to be like a full picture right let me reset set the skeletone again looks more elegant and then we can try again and Define more things in the text prompt let's say sunny day oh yeah by the way I forgot to add some Laura's models in here I should do that early I should do add details and add a flower Fashions clothing for this one and then I put a little bit higher for sampling steps and here is the result but the face is kind of awkward we can fine-tune that later so let's move the pose again on the vector layer and like a person sitting or bending down in a garden and add ultra realistic some more text prompt cooking it okay let's enhance the character's face and also add some details on it okay the face still looks a little weird and let's see if we can change the sampling method here to DPM plus plus S2 and let's see that will be something a little better yep there you go okay okay something better than before and this is also the first time I tried with this photo editor and I'm still exploring this new stuff and you can also upscale the image right on top of the drop- down menu and then I select the ultra sharp upscaler and after that we have a new generate upscaler image and here's the result well looks better actually the upscaler also do a refiner feature as the checkbox enabled by default here so the face and everything it looks more better improved the image quality yes so try it out guys that is the Crea image editor with stable diffusion KFA UI it's open source totally free we can use that with stable diffusions so let's see let me explore the LCM real time drawing on another video for stable diffusion AI images with image editor it can be level of enhancement for images so have fun guys and I will see you guys in the next videos have a nice day see you

Channel: Future Thinker @Benji

Views: 3,388

Keywords: Stable Diffusions, Krita photo editor, Open-source photo editing, AI image generation, Comfy UI integration, LCM model, GitHub installation, Python plugin manager, Control Net Open Pose, Upscaling images, Real-time drawing, Image refinement, Free photo editor, Image enhancement, AI image generation tutorial, Image editing workflow, Image prompts, Krita AI Diffusions plugin, Comfy UI manager, Image upscaling, Community engagement, Image refinement tutorial

Id: umb4ddC3gK8

Length: 10min 59sec (659 seconds)

Published: Thu Dec 28 2023