Tortured women

wulf · Post by **wulf** » Sun Jan 25, 2026 1:32 pm

Arcas wrote: ↑Sat Jan 24, 2026 8:57 pm I also posted two today because I now have an inferiority complex thanks to all the cool stuff you can do with AI. I love seeing my original come to more realistic life, but it leaves me feeling.... like a Poser artist

I understand that you might feel that way, but remember that AI can't do the base image (yours) anywhere near as good as yours from scratch... at least not yet. Truthfully, your art doesn't need improving imo.

Arcas · Post by **Arcas** » Sun Jan 25, 2026 1:40 pm

wulf wrote: ↑Sun Jan 25, 2026 1:32 pm I understand that you might feel that way, but remember that AI can't do the base image (yours) anywhere near as good as yours from scratch... at least not yet. Truthfully, your art doesn't need improving imo.

Thanks for the sympathetic understanding.

For what it's worth, I've been looking at my older works and comparing them with what I do now, and at least on the technical side I feel like my skills have grown since I semi-retired in 2016.

xs70 · Post by **xs70** » Sun Jan 25, 2026 5:56 pm

Arcas wrote: ↑Sun Jan 25, 2026 11:42 am Thanks for your feedback on the latest two images. IIRC, both had a bit of Bloom enhancement, but not much. They were both kinda rare in that I looked at the raw render and didn't immediately see a laundry list of postwork that needed doing. In the floor piece it was mostly just the hair that got the AI boost. And I don't think I used Bloom at all on the carry pic. So maybe I'm finally figuring out what I'm doing?!!

Ah yes, that makes sense. It's well integrated, I couldn't even notice.

I'm someone who doesn't subscribe to the 3D vs AI paradigm. I'm sure you remember when Photoshop was new and "real" artists were complaining that digital art wasn't real art, Photoshop was only good for bevels and lens flares (and there was a lot of that), etc. We're now in that same phase with AI, just 20 years later. It will pass. To me, AI is just another tool in the repertoire. The question isn't if one should use it. If the result looks better after using it, then use it. If not, then don't. I focus on the end result. I really don't see why people would not want to use AI, if it's well done. Or why people would only want to use AI. It's missing the point, it's the result that counts.

For the images you make, essentially cinematic torture scenes, the best pipeline today is 3D -> AI finish. Like you're doing. 3D to set up the scene, AI to finish the people in the scene. The human body, or faces/facial expressions. Only if necessary, and it has to be well integrated. In a couple of years, everyone will be doing it that way. It's just a better overall result.

In my own images, my camera position tends to be close to the girl. Closer than in 3D images. I'm showcasing beautiful girls first and foremost. My viewer is on the first row looking at the girl, too close to not be considered a participant, often the torturer himself. I'm essentially making Playboy shoots of the prettiest girls I can make and placing them in GIMP scenes. So I need more human detail in my images since I'm closer up, like detailed nipples, pretty faces, subtle facial expressions; and less cinematic scene around the girl. So I have less use for an initial 3D scene, to the point where I don't need an initial 3D phase.

Arcas wrote: ↑Sun Jan 25, 2026 11:42 am As for two images per day, I might do that again, tomorrow... but that leaves one image left in the Day One sequence. After that I'm thinking of changing my approach, entirely. Doing this whole thing as a coherent, sequential story kinda sabotages my usual inspiration method and proven strengths. So after this, I think I'll just be doing a series of Videodrome one-offs, without trying to rationalize where and how they fit. I think that'll help me deliver the kind of stuff folks wanna see... and let me create at my sadistic best.

Sounds good to me.

While we're showcasing video, here's a couple I made a couple of months ago. Dunno if you've seen these, they're a series. I've stopped doing video for now, I prefer stills for some reason. But I was kinda happy with these.

https://motherless.com/AFACB6B
https://motherless.com/2073F2C

Ed2 · Post by **Ed2** » Sun Jan 25, 2026 6:43 pm

thx1139a wrote: ↑Sat Jan 24, 2026 5:30 pm Ask and thou shall receive (for stuff like this as much as possible, anyway)...
Five versions... Experimented a little... adding speech usually isn't great but I thought two attempts were okay... one is an attempt to channel Paula Warner's highly stimulating "Please don't hurt meeeeeeee...." from Don't Answer the Phone. As with all things AI, you never know what you're going to get exactly and one version has some interesting stretching involved. But, without further ado... https://ufile.io/7hy4d5g5

Lovely animation and vocalizations. I look forward to seeing her react to other implements.

Ed2 · Post by **Ed2** » Sun Jan 25, 2026 7:04 pm

wulf wrote: ↑Fri Jan 23, 2026 1:33 pm As a collector, I suggest leaving Arcas's name on them as you did, but also add your signature, or indicate it's AI and in some way who created it (with Arcas's permission of course).

xs70 wrote:The problem is I don't have a signature, I don't sign my images. So it's a bit of a conundrum

I also enjoy images without obvious signatures. I think they distract from the action at best and compete with it at worst. They are almost as problematic as watermarks. Another artist I greatly admire also refuses to sign, Williej, and I believe his images are the better for it.