- Jul 21, 2018
- 3,733
- 61,489
That is true lol. I will just jump over to other threads that haven't had much going on in them.Thank god for the ignore feature. I won't see any of it more than once from the same person.
That is true lol. I will just jump over to other threads that haven't had much going on in them.Thank god for the ignore feature. I won't see any of it more than once from the same person.
Thanks one more question how you generate request to receive pictures like that? Because I receive some shitty things from that ai)Stable diffusion
AI advances very fast and good results depend on knowing how to create promps, how to train and use LoRAs, and checkpoints, on knowing how to use Controlnet to improve postures and avoid deformations. In short, it depends on the results you are looking for, you can do something very basic and quick or you can dedicate time and work to create something very specific with the details you are looking forThanks one more question how you generate request to receive pictures like that? Because I receive some shitty things from that ai)
let's wait to see what Stable Diff SDXL can doAI advances very fast and good results depend on knowing how to create promps, how to train and use LoRAs, and checkpoints, on knowing how to use Controlnet to improve postures and avoid deformations. In short, it depends on the results you are looking for, you can do something very basic and quick or you can dedicate time and work to create something very specific with the details you are looking for
You must be registered to see the links
the photo does not open, writes" Error when establishing a secure connection
I like this one
I save it for the "AI is the future, bro" bros. Invariably they are the same people who were saying the same thing about NFTs last week.Every time I see that crap, I add that person to my ignore list.
I have nothing against messing around with AI... I have done it a LITTLE myself. But... I think in here, if you're working on the sort of graphics we have in visual novels like renpy, stick with DAZ. There's a time and place for it, this isn't it. At the very least, stick to the forums on here that have been created for it. I come to this page less and less as I get tired of seeing it and photoshopped garbage. The assets for this game have been posted, I posted some of my own scene subsets to help get people started. There's no excuse for not having at least some basic DAZ renders. Nobody will criticize you if they're not perfect. I love seeing people learn to use it and watch their work grow.I save it for the "AI is the future, bro" bros. Invariably they are the same people who were saying the same thing about NFTs last week.
The thing about Daz is that when you get your character dialed in, you have an actual 3d model that you can bring into Unreal etc. With AI we're gonna be waiting for a while to get the amount of fingers right in a 2 dimensional picture.
I've never been a crypto guy. Ever. I laughed st NFTs. Etc.I save it for the "AI is the future, bro" bros. Invariably they are the same people who were saying the same thing about NFTs last week.
The thing about Daz is that when you get your character dialed in, you have an actual 3d model that you can bring into Unreal etc. With AI we're gonna be waiting for a while to get the amount of fingers right in a 2 dimensional picture.
I love marks this is a hot as fuck art
Exactly. Right now the difference between something that sucks and something that is shockingly realistic/good is the way you communicate with the software via prompts. That and what training model you use.AI advances very fast and good results depend on knowing how to create promps, how to train and use LoRAs, and checkpoints, on knowing how to use Controlnet to improve postures and avoid deformations. In short, it depends on the results you are looking for, you can do something very basic and quick or you can dedicate time and work to create something very specific with the details you are looking for
You must be registered to see the links
Future is here) thank you . I’ll try to figure out with that AIAI advances very fast and good results depend on knowing how to create promps, how to train and use LoRAs, and checkpoints, on knowing how to use Controlnet to improve postures and avoid deformations. In short, it depends on the results you are looking for, you can do something very basic and quick or you can dedicate time and work to create something very specific with the details you are looking for
You must be registered to see the links