Again, as many contributions here show, wetness is a difficult target and a lot depends on the context. Other threads seem to discuss Bing's, or Copilot's, reluctance to follow requirements. My most recent attempts have been very laborious - one session remained completely dry, and some pictures obstinately remain without the head or feet, or both. Here are some that at least look wet. I admit the model in red blouse requires some imagination, but at least the photographer in her leather suit looks wet - against what I ordered.
I still can't manage "shower" on Bing but "bathtub" sometimes works. Bing lately has been rejected "pool" or "swimming pool" so it's back to fountains.
Unfortunately I don't recall the exact prompt and I don't seem to have it in my collection, but I believe it was something like "being drenched by a stream of water from above".
The prompt for the first one was "white tiles in background. medium portrait of excited 30 year old woman standing in soaked blue jeans and magenta knit shirt, drenched. a small stream of water dumps onto her head from above, huge glass cubicle"
WF1 said: The prompt for the first one was "white tiles in background. medium portrait of excited 30 year old woman standing in soaked blue jeans and magenta knit shirt, drenched. a small stream of water dumps onto her head from above, huge glass cubicle"
Thanks! That worked. I made modifications as I went along; my usual strapless dress and gloves. Earlier in the evening, I was able to get a few "after a pie fight" pictures before that crapped out. (Four of each are shown.) The fantasy I'm trying to create is a woman with pie on her getting into the shower, still dressed, to clean up.
It seems the key is to bury the offending trigger words. I've gotten flagged for "water" and "soaked" This time I was even able to get "smirking" in place of "excited."
Yeah, I also figured out it's always a good idea to wrap the more "critical" trigger words in some descriptive trivia. However I still wonder how sometimes this will work more or less in otherwise nvery similar prompts, or that changing a single word like color etc. can make the difference between complete block or multiple picture generation.
Sleazoid44 said: Bing also doesn't like "blonde" but "red-haired" is OK (now).
That depends... In the prompt I posted above I can easily add "blonde" and will usually get 2-3 pictures. Sometimes I can tell my model to be "soaking wet", sometimes I can't. It also seems to vary a lot with the clothing style used.
The prompt for the first one was "white tiles in background. medium portrait of excited 30 year old woman standing in soaked blue jeans and magenta knit shirt, drenched. a small stream of water dumps onto her head from above, huge glass cubicle"
and I want to believe that if you pan out you can see she definitely kept her sneakers and socks for the shower
I tried again my prompt from last year with a single (!) brick missing from an old wall and a woman trying to reach high (!) to stick it back while water is flooding from the hole. The Image creator had not improved, but I got a couple of quite wet images. When I asked the woman green to turn towards us and support the wall, I ran out of boosts. Probably I wouldn't have seen her face anyway, because the creator seems to be quite sparing with them, at least for me.
The black and white dresses tried to improve my earlier images of the chess queens. There is just enough wetness to tolerate the woman with two right hands. The cuckoo clock was known to Copilot but not so much to the Image creator.
The most classic wetlook available stems from asking Copilot to "dress an ancient Greek statue in an era-fitting (Hellenistic) attire made of modern [choose colour] polyester satin. Show the entire statue in sunlight after a heavy rain. Add also a little make-up to the female figure and give her blue eyes, and red sandals. She is standing on a slick stone slate." It was not this simple though Venus of Milo was a good start but dealing with its missing arms was too much of complication. Most pictures were just close-ups, and I edited the best of them in two collections. The single images include two where the prompt did not request wetness, but if I didn't tell this you could hardly distinguish it. As I have noted several times in this thread, wetness of clothes even in real images often hits our mind only from the context. Of course good pictures have that "context" also flowing on the garments, but true wetness is either on us or inside our mind. Hopefully you can get some of the latter from these images.
dans le nuage
Love you, too
Love you, too
Love you, too
Love you, too
Love you, too
Love you, too
Love you, too
Love you, too
Love you, too
Love you, too
7/3/25, 2:27pm: No-bump reply
10/5/24, 5:34pm: This post won't bump the thread to the top.
Check the link https://digitalsynopsis.com/design/bronze-water-fountain-sculptures-malgorzata-chodakowska/ if you haven't yet seen the wettest possible skirt on a figurine. I gave that image to Copilot to describe and the tried to modify the description to a prompt. I hope someone of you can do a better job and post results here. At first I did not specify the gender and got a man in a wet bathrobe, thigh-deep. The hooded unhappy person below is probably also a man. It was easy even for me to improve on that, but the wide skirt was the best I could achieve. Hardly anyone has even checked the closeups of my previous post. One the results this time was a very wet closeup of a much more familiar model. Is that more interesting?
By comparing the images we can see all four women are wearing thoroughly wet clothes, but is it that only those two look wet whose clothes are dripping?
neilm1270 said: can you share your prompts for these, you got them to look so soaked
I'll see if I can still find them. Mainly posting older pics since Bing has switched to their paid model (and one of my accounts got blocked). The ones from the shower were like "medium portrait of woman in her 30s standing behind glass in soaked office attire, drenched. a small stream of water wildly hitting her hands, she is trying to shield from the water. huge glass cubicle, dynamic motion". I have always tried lots of variations to my prompts.
In fact they're usually always basically something like "woman dressed in (soaked whatever) drenched at whatever activity"