Video

Could make bank creating stickers for etsy with this technique.

what GPU are you using for this?

I disagree with the sentiment that we should 'embrace AI'. AI cannot and should not produce art. Art is one of the most important aspects of life and it shouldn't be diluted with artificial and substance-less illustration. Not only will jobs be lost to AI but so will an aspect of humanity and all for what. Why? Why should art be 'easier'? Why remove the process from art? Why exist if we can be stripped of purpose by lines of code?

This helped me get started thanks! had to use 3.10 tho - 3.12 had errors

It is indeed very useful addon, thanks!

The "Blender Startup File" link doesn't seem to work. It says: "This content is no longer available."

Inpaint also doesn't work well with layer diffuse.

TypeError: 'NoneType' object is not iterable

why not inpaint instead of photoshop?

Is it possible to load a lora into only the foreground and another lora for the background? Or maybe two loras for the foreground and only a prompt for the background?

wow!

Thank you so much. this was amazing! i did subscribed and liked.

Nice to have you back! Huge post! Wow! I know what I’m doing tomorrow. 😅

RIP AMD users like myself.. I got A1111 with directml working but havent figured out how to get anything else to work. This included. I have a 5700XT AMD card... no rocm support(whatever that means) wish I had the money for a basic nvidia gpu

Fascinating, great video!

Awesome!

Nice to see the best SD CZcamsr is still at it.

Welcome baaaaaaack ! I was waiting for you realy !

Great video!

Is there Anything like this for foocus?

Awesome

i dont have DPM++ 2M Karras available

Good work, and all with good description step by step, with free tools

Great tutorial as always. I have a question though. This is great for materials but... what about images representing real things? I mean, Imaging I want to create the texture that goes in a circular surface representing some concentric circular stones with symbols. I have been able to generate a texture for this (an image in stable diffusion) and apply it with blender. The problem... it is lit, and I want it to do the same process you have followed but for this kind of textures. Create albedo, normal, etc. Is it possible? Or this models only generate "materials"?

I'm really curious about the AITA in the third browser tab

At first everything is fine but when I hit generate it takes forever to create an image, the blue bar almost does not advance, I have a ryzen and a 6gb nvidia gtx 1060 video card 😭

Excellent.

If you didn't find the tiling option active, remember to go to Settings -> User Interface -> Settings in UI, add tiling to Settings for txt2img and if you need, for img2img too

I've never done video game texturing before, so maybe I'm wrong about generating water textures, but this just generates stones, dirts and bricks?! I've played enough video games before to notice that a lot of them look very similar if not identical. All beautiful. Does water in games have textures at all, or does this model simply lack images of water to properly generate it? Edit: Oh, and thank you for this video! This has helped me!

thanks for the explanation

It's the biggest piece of dog shit, fully of bugs, erros, fails, crashes, a total mess addon.

I'm massively struggling with this. I cannot get Pytorch to recognise my GPU, despite confirming that it's compatible and checking I have the correct CUDA version that's compatible with Pytorch. I've tried reinstalling Pytorch, installing Nvidia developer tools so I have the correct CUDA... nothing works. Pytorch just keeps telling me CUDA isn't available. torch.cuda.is_available returns False and torch.version.cuda as None despite installing a CUDA version of Pytorch. I've no idea why it's been so easy for others!

Does it really makes sense to do the liquify step BEFORE the img2img-generation? Since the exact position of the eyes and other details is really only influenced by ControlNet. (It would make sense to do this in post though, after the final generation.)

Awesome tutorial.

So this can’t be done on a macbook m1?

Shows why you actually have to be an artist and understand perspective and understand other elements to make it more dynamic. No one using Ai is going to make the kind of money a real artist makes. You guys are trying to reach for the coin between the couch cushions; the more you reach for it, the further it gets.

How do I restart this program in my browser?

cool video

This add-on is so buggy. I could barely get it to work and even when I did - the generated texture did not follow the uv at all. I don't see how it's using the depth here.

You're the reason I started my Stable Diffusion journey around a year ago! My life has changed for the better since then. genuinely, if I had money, I'd send some your way to show real appreciation.

thank you so much for this tutorial!

is there any problem on Python 3.11? I done all task get error 128 on run bat and tried many solutions not working

holy sht git gives me so many gawd damm questions when installing its almost feels like a joke

exit code: 1 stderr: C:\Users\62877\AppData\Local\Programs\Python\Python310\python.exe: No module named pip Launch unsuccessful. Exiting. Press any key to continue . . . when clicling on web-ui user.batch , it showed that notif, please help😮‍💨

This was cool. Thanks for sharing

How do you get it to run so fast? I installed ControlNet and it takes about 20min per image now

stable diffusion is the most fun I've had in awhile.. a steep learning curve but great hobby. As a beginner, the most confusing part is trying to figure out what files go where and why... such as how do you know when a model requires an additional download that goes in another folder.. what is the purpose of a yaml file, etc. Someone could get a lot of views if they could go in depth on demystifying SD 🙂

I'd rather just paint it from scratch.

if ai is used as a tool and everyone keeps they're job then there's nothing to be afraid of but if everything is ai in the future then we are fucked, and it's so sad to see so many people with so little hope about the future.