This is a Z Image LoRA to create persons bound with their arms up, above their head, usually tied with ropes. The LoRA supports the following:
Persons bound with their arms above their head, wrists bound together (hands crossed or in parallel).
Rope hanging from above and to the ropes around the wrists.
A gag (or not). It can support a ball gag (any color), a tape gag and a cleave gag (see samples, the cleave gag takes a couple of efforts more)
Bound legs together, or to something else (railings or whatever), or legs spread (and bound to something at the ankles)
Several views (front/side/angle/back view)
A person placing his hands over the bound person.
Description
Z Image Base port
FAQ
Comments (3)
Thanks for working on all these models.
(seems like you'll be the Joschek of this generation's frontier models, eh?)
I'm working on restoring a yellow buzz war chest, eventually I'll be able to afford entering the frontier myself (genning, not lora training). But judging by the showcase pics on a few of your recent models, it seems like ZI base isn't significantly better than ZIT, right?
(at least for bondage. i.e. Sometimes (but not always) added detail, yeah, but often at the expense of blurring, and that detail is often irrelevant and at expense of main subject. And imo the poses are often less sexy; more robotic and stiff).
Do you have a better assessment?
I think that's also a matter of personal taste. ZIT is a quite good compromise, and in just a few steps you can produce a nice image. ZIB is supposed to be better, but you need to prompt more to achieve better results. I can't call my self an expert on any of the two, although I 've that ZIB can pull out more complex scenes
@zkrd Thanks. I hadn't heard that ZIB needs more detailed prompting to achieve its superior results over ZIB. Good to know, I'll have to research that (BTW, you bumped into any good ZIB usage guides?).
And yeah , it does make sense the the pose stiffness etc I was complaining about is probably my personal tastes (messing me up, as they do).