KFproduction
patreon
Sorry, been a bit distracted
🕑 Added 2022-10-28 22:56:22 +0000 UTCI recently discovered this new wonder to me, maybe you have run across it in the past month or so, a thing called Stable Diffusion GUI, which lets you basically take text prompts and turn that description into an image..well, sort of. It's only as good as the prompt itself..and of course the models the AI has to draw from..it takes up about 7gb of my main drive .I am an infant in this strange universe and trying to learn as quickly as possible, but as frustrating as it has been, it is quite addicting and once in a while I get something I actually like. My goal would be to use 'image to image' results instead of 'text to image' and hopefully enhance my rendered work in some way. Right now I am just trying to get the basics down and having a lot of fun. I will cherry pick and share some of my results if you guys are interested in seeing what it has produced so far. Let me know if so.
Comments
I agree, wasn't trying this out as a solution to replace my use of Daz by any means. But from seeing how far AI has come since it couldn't even create a stick figure cow that you could recognize, I just had to try it out and see what I could prompt it to make. It is still quite limited but also exciting to see how quickly it is evolving.
As of yet, I don't think this Stable Diffusion and other such "AI" 3D rendering programs would surpass any other commercial 3D programs. The results look weird to me. The program you're currently using with all your art is superior.
FomalhautBravo
1.6? I thought they just made a 1.4 like days ago...and, I'm not jealousing about that 3090, no, not a bit. Really. (And it really feels better to have conversation on discord)
Well, given I am banned from dA I have no clue what is going on there. I have seen a little of it on some discord channels, but I got into it at v1.6 and it has been turning out better as i go, trying to pick up tips along the way. Helps that I have a RTX 3090 to support the requirements to use up-scaling and regen functions.
FomalhautBravo
I thought that you've already heard of this stuff that has sweeped the entire DA:D I thought most people out there are using Midjourny and Stable Diffusion, yet I'm still doubting about the result, recent 1.4 models is...not better than 1.3 or earlier models, as far as I feel
