(KRITA AI) INTRO to Stable Diffusion for Krita PART 2
Вставка
- Опубліковано 1 сер 2024
- This is part two of my complete introduction to the Generative AI for Krita plugin where I go over the tools and features.
Resources:
Krita: krita.org/
Plugin: github.com/Acly/krita-ai-diff...
Installation Instructions: www.interstice.cloud/plugin
Required Models and Nodes: github.com/Acly/krita-ai-diff...
CivitAI: civitai.com/
Music by CreatorMix.com
Extremely easy to follow! we are waiting for more videos like this!
Thanks! I'm looking forward to doing more!
Muchas gracias por compartir muy buen trabajo
Thank you!
Thank you
Almost missed this. Thanks for part 2.
thank you so much! but i have a problem, i don't get the live or animation drop down menu :( like they just don't show up, only generate and upscale.. i don't know why and don't know how to fix it...
That's really odd. Maybe try reinstalling the plugin and make sure both the plugin and krita are the latest version.
is there any way to paint a image roughly and ai complete / finish the work....?
Yes, try starting a painting and setting the strength to about 55% to 65% in the Image Generation or Live Mode.
Any idea how to use other model types on Krita, like the upscaler, Lora's using weights, and Lycoris?
You should be able to add models by adding them to the appropriate directories in the ComfyUI installation you are connecting to. You can add loras directly in the text prompt or in the settings. Lycoris should work the same, but I'm not sure. github.com/Acly/krita-ai-diffusion/releases/tag/v1.10.0
@@IntelligentImage-sl7uf OIC, that answers my question. Thank you so much. I can Finally and use Lora weights now.
Glad I could help!
Started playing with this last night. Straight forward tut! Thanks! Surprised I might want to use Ai after all! Modifying my artwork seems internally plausible since I composite like crazy... modify not totally create via prompts....for concept art might start using this workflow
Thanks! I'm working on a workflow for using AI only at the concepting stages. I think that may be the purest use of AI in digital painting.
How do I uninstall the plugin?
I'm trying to update it from 1.16 to 1.17.2
You don't have to uninstall the old one. Just install the updated version the same way you did the previous version and it will overwrite the old one. There is more information here: github.com/Acly/krita-ai-diffusion/wiki/Common-Issues
Hey man, I love these videos! I got some quick questions if you don't mind me asking,
1. What model are you using in this video? It looks really nice.
2. When I use a LoRA in Krita AI for some reason it always keeps the strength at 100% no matter what I do, do you know anything about that?
3. What brushes did you use in this video if possible?
Thanks! I used "Store Bought Gyoza" for the model with some "Epic Realism Natural Sin" mixed in.
I'm not sure about the LoRA strength issue. Do you change it and it changes back to 100% or are you unable to change it at all?
I just use the stock Krita brushes. I use the "Charcoal Rock Soft" brush almost exclusively for painting, but I am trying to find others I like to add some variety. I have just gotten used to that because it is good for both blending and carving out shapes. I also use the airbrush for blending of course.
@@IntelligentImage-sl7uf Thank you! I change the lora strength and after I apply all the settings it reverts back to 100%, really weird behavior but I haven't seen anyone have that issue online after some googling
Are you using the most up to date version of the plugin 1.17.1?
@@IntelligentImage-sl7uf yup
You could try updating to 1.17.2 now if you haven't already and see if that fixes any issues.
Hello, live and classic mode are very slow on my setup, would you know the minimum gpu requirement for that ? I'm on a gtx1080 which works well witg Automatic111 and ComyUI outside of krita. Thanks.
At least 6GB VRAM is recommended. Check your canvas size. Try 768px x 768px and go larger from there if that makes it run faster.
@@IntelligentImage-sl7uf my canvas is A4!!! pretty sure it's rhe reason why big thanks to you
Glad to help!
@IntelligentImage-sl7uf do you know what recommended canvas sizes are in general vs someone with a 4090. I am trying to figure this out but am not really getting anything. I have an outline of a female but when I generate it just looks like trash. Is it because my canvas is to small or is something else messed up. Hmmm.
@@dylananhorn1 You should be able to get something decent looking at 768x768px. 1024x1024 would be better if your system can support it. It could be a different issue though. What model are you using?
How to install the plugin on mac
There are Mac specific instructions under the "getting started" section of the download page: github.com/Acly/krita-ai-diffusion
Can you make comic books and other sequential art with this?
Character and background consistency would be the hardest thing right now, but it is definitely doable. It's what I want to do in the future.
Same, I really want to make my own comics and animations but I can't draw very well and I tried to work with artists in the past and it was a nightmare and very costly. I don't know Daz or Blender, either but maybe AI art could be a good solution. I will probably buy a new computer next year, too, so it would be a great time to test out hopefully a better version of this AI come 2025. Cheers. @@IntelligentImage-sl7uf
github page not found
Fixed it! Sorry, apparently when UA-cam pulls URLs from an older video, sometimes it doesn't copy the whole thing. Thanks for letting me know.
Please step by step instuctions
Good idea. I'll do a video on a workflow.
It takes time to generate the image
Depending on the hardware you are using, it might be slow. Try using a smaller canvas size. Try 512x512px and see if that speeds things up. If that is usable, try a little larger.
It really sucks that this is comfy ui only... no a1111
If you let the plugin install its local managed server, it will run everything for you and you shouldn't have to actually set up or use ComfyUI if you don't want to.
@@IntelligentImage-sl7uf strange. i tried a local a1111 server but the plugin doesnt even show up in Krita to enable. As far as i know there theres no such thing as a live preview with the a1111 ui
I hope it can do non-perverted images too.
noooo
It should be possible, but I haven't tried it.
@@IntelligentImage-sl7ufMega Based 😂