@Background Pony #E129
I don’t even know how to crop image properly and I am fairly certain that many people using AI for art generation aren’t the best at this stuff themselves :3
Imported from Derpibooru - Posted by FluttershyIsNOT aDoormat
@Background Pony #E129
Some people don’t even know how to paint little bits. They might not have input hardware that can easily do it like us.
Also moving the torso to fit properly, would mean having to paint more of the chair. Still easy for me, the chair is not remarkably detailed.
I haven’t tried generating any of my own images. I’m sleeping on it, until it gets really cheap and impressive. I still got traditional digital art commissions to work on.
Also going to build my own AI machine, and feed it with stock images and my own art. The way nonartist feed these AI is kinda gross. These tits aren’t art, it’s undisciplined animal impulses.
@Background Pony #E129
I am a complete art noob and even I know how to clean up anomalies left behind by stable diffusion.
A little effort in inpainting is all that is needed.
This is a feature that is usually available in the same software you use to generate the base image itself.
Sure I have been experimenting with SD for months now but this is still nothing compared to actually being an art pro so in the end I would argue that anyone with a little bit of practice can produce pieces that do not necessarily look like the above.
@Background Pony #E129
Assuming they have the skill. Not dumping on them, I’m just saying it’s easy for me to do because I’ve done art for over 10 years, not so easy for everyone. Also I’m only fixing AI art for money, since it directly competes with digital artists.
EDIT: Ye i know soon they won’t need anyone to fix their work. It is what it is.
That’s so easy to fix. Simple editing skills could take away the top hand and move over the body a little (the chair being a little weird is less noticable than her waist being very weird.)
I don’t even know how to crop image properly and I am fairly certain that many people using AI for art generation aren’t the best at this stuff themselves :3
Some people don’t even know how to paint little bits. They might not have input hardware that can easily do it like us.
I am a complete art noob and even I know how to clean up anomalies left behind by stable diffusion.
A little effort in inpainting is all that is needed.
This is a feature that is usually available in the same software you use to generate the base image itself.
Sure I have been experimenting with SD for months now but this is still nothing compared to actually being an art pro so in the end I would argue that anyone with a little bit of practice can produce pieces that do not necessarily look like the above.
Assuming they have the skill. Not dumping on them, I’m just saying it’s easy for me to do because I’ve done art for over 10 years, not so easy for everyone. Also I’m only fixing AI art for money, since it directly competes with digital artists.
Usually it only takes 5-10 minutes.
yeah, I guess it was to be expected to happen sooner or later
It’s garbage AI art, which often can’t figure out how hands work.