-
Willkommen beim WoH Board. Um Beiträge zu verfassen, musst du dich kostenlos registrieren.
[Sammelthread] stapfschuh's new home for sharing prompts
- Ersteller DavidBowie
- Erstellt am
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
just something quick:
won't test any prompt on xl anymore. reason is just that those checkpoints all have a different understanding of prompts and testing every prompt and finding a solution for all of them is just way too time consuming. standard sd simply has the best understanding of prompts and ideas. it may not generate the best anatomy, but i have solutions for that.
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
won't hide this one, because it's just random:
something similar like greta thunberg and communication happened right now. i prompted in negative "fake-photography" and got this
can you find out what my checkpoint is trying to tell me? if not, then look at her and why my checkpoint generated this close-up on purpose and why it does look like a character from some 3D game? my checkpoint just wants to tell me that it can't generate any realistic images because there is 3D mixed in. especially "3D-ingame-graphics". and so i put that in negative and i was 100% right, because the next image that got generated after this one above and with that new negative prompt now is this one
as you can see, "images have their own language" (my grandmother once told me the same when i was a child xD) and your checkpoints know how to use it but you have to listen very well sometimes.
Post automatically merged:
and this one is similar to that last post. a pattern that i've already seen dozens of times already that was super annoying. images like this
all i prompted in positive and negative was this
(((Ultra-HD, reallife-reality))) 1girl, outside
(((simultaneous, made-in-china, 3D-ingame-graphics, frail)))
nothing here that says "first-person". but that's not the point. the point is what that girl is doing right now and where and why something similar could happen.
she's adjusting her hair. correct, but this one, "adjusting-hair", only ended in freaky hands on girls heads in the past.
let's just think about "why" she's adjusting her hair and "what" is needed to see "how" one's hair would look like. and don't think about the "viewer" being a person only.
"why", "what", how" and "viewer... the answer is a mirror. you, the viewer, are actually a mirror at those moments. let's test it prompting "mirror-view" in negative
and i was right. i could upload more, but those all are the same....
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet:
DavidBowie
Ordensbruder
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Post automatically merged:
Hidden content
Zuletzt bearbeitet: