A New day, a new beginning

[Theo]

Slightly different style for today's selection, hope you like them.

I do like this style

[Jim_Weathers]

Regarding the prompts, I usually put "water" in my negative prompt...

I finally learned this a while back as well. The next problem was constantly fiddling with the weight to give it based on what looks good versus personal experience. It's not the same thing. It also varies with different models. Can't wait for AI to get smarter. Beautiful work as always.

Thanks. Weights are magic, you get a series of ankle deep that you can't shift, then suddenly its up to the armpits and you wanted something in between. One thing I've found is that if you specify footwear, AIs always seem to think you want to see it, so either the subject won't sink, you get extraneous limbs or you get shoes/boots on hands etc. The other thing that bugs me is trying to get some arm movement, it's really difficult to get the arms away from the side without it going over the top! One other thing I usually put in the negative prompt is "mouth open", I find I keep getting goldfish expressions if I don't do that!

I was looking for the exact opposite - but then again, in the pics I did, I was going for a more dramatic look. I’ve noticed if you don’t, the I decides to put that disinterested model look. But to each his own. And yes, I learned almost immediate;y to put water in the Negative prompt either than or you just hate a bunch of lake and river pics

[cerberus]

Yummy, today some innocent mudplay out at low-tide!

Very nice pics, as long as you don't look at their hands. 6-fingers are back

Aggggh! Slipped through the net, I usually count the fingers etc, but I was a bit late and rushed. Oh well, thanks anyway.

Very nice pics, as long as you don't look at their hands. 6-fingers are back

Hey, don't knock it till you try it. Polydacs give the best hand jobs!
I'm sorry... I couldn't help myself.

I couldn't possible comment on that!

Slightly different style for today's selection, hope you like them.

I do like this style

Thanks, they were a bit of a struggle, too many twins and limb piles!

[cerberus]

Regarding the prompts, I usually put "water" in my negative prompt...

I finally learned this a while back as well. The next problem was constantly fiddling with the weight to give it based on what looks good versus personal experience. It's not the same thing. It also varies with different models. Can't wait for AI to get smarter. Beautiful work as always.

Thanks. Weights are magic, you get a series of ankle deep that you can't shift, then suddenly its up to the armpits and you wanted something in between. One thing I've found is that if you specify footwear, AIs always seem to think you want to see it, so either the subject won't sink, you get extraneous limbs or you get shoes/boots on hands etc. The other thing that bugs me is trying to get some arm movement, it's really difficult to get the arms away from the side without it going over the top! One other thing I usually put in the negative prompt is "mouth open", I find I keep getting goldfish expressions if I don't do that!

I was looking for the exact opposite - but then again, in the pics I did, I was going for a more dramatic look. I’ve noticed if you don’t, the I decides to put that disinterested model look. But to each his own. And yes, I learned almost immediate;y to put water in the Negative prompt either than or you just hate a bunch of lake and river pics

Yeah, getting the right expressions can be a struggle sometimes, the AIs don't always respond to instructions. Generally, I prefer to have them looking away from the "camera", to the side, up, down, etc this way, at least they look like they are concentrating on what they are doing and aren't obviously aware of the "camera". If they are looking at the camera, then you want the eyes have to be showing some emotion of some kind.

[cerberus]

Well, today I concentrated on two things, one was getting realistic looking cold weather fishing type gear, it all seems too thin really and SD hasn't got a clue what "fishing waders" are, so I tried various permutations and came up with "(thick dark neoprene fishing waders:1.4)". It was OK originally without "dark", but then for some reason they all went light. Varying the number before the last parenthesis increases/decreases the thickness (if you overdo it you'll get the sort of stuff people who defuse bombs wear). The other was to try to find a solution to the "twin" problem on wide images. I'd read that SD was trained on square images of 512x512 dimensions, so I reasoned that it would treat a wide image (say 768 wide by 512 high) as an image plus a part image and do a sort of merge resulting twins, joined bodies and extraneous limbs. To test this I set SD to create 2560x1440 images by generating a base image of 640 by 360 (hoping this would be close enough to 512x512 to prevent a "twin"), then using the upscaler with a factor of 4 to get to the full size image. Surprisingly, this seemed to work, as you will see if you look at the imagebam folder at the end of this email.

00417-3326427948.png

00430-1079549650.png

00437-2500942949.png

00439-2726826465.png

Watch this last one, if she invites you round for a meal and there are fava beans and a nice Chianti run like crazy!

00441-2103475164.png

So, this girl is talking to her friend and she says, she's getting fed up with her boyfriend, he's a bit of a nasty bastard and a mud/sinking fetishist as well, which she can't stand. Anyway, she's found somebody she likes better. She's meeting him one last time at one of his favourite remote sinking spots, then afterwards she is going to break up with him. Her friend asks if she is sure he doesn't know about her new love interest, but she says "No he's too dumb to notice". Well what could possibly go wrong?

https://www.imagebam.com/view/GAAB06

[bogbud]

Have never seen that type of fishing gear, but I very much like what the AI made out of it. Should be produced for real!

I particularly like this one attached. She could be really wearing proper chest-waders (already swamped of course). But that area does look more like a clay pit or a settling pond at some extraction-site. Not the area for fishing. So she may be a worker there or a fellow like-minded, sinking-addicted gal. I hope for the latter!

[sixgunzloaded]

Watch this last one, if she invites you round for a meal and there are fava beans and a nice Chianti run like crazy!

Yikes, the crazy eyes..! The teeth..! The...the...chest thingies! Are they outside her catsuit, or are they bursting forth from within???

Did the AI do it like this on its own or did you prompt it that way?
If it's the former, then maybe it's time to be afraid.
Still lovely to look at though!

[cerberus]

Have never seen that type of fishing gear, but I very much like what the AI made out of it. Should be produced for real!

I particularly like this one attached. She could be really wearing proper chest-waders (already swamped of course). But that area does look more like a clay pit or a settling pond at some extraction-site. Not the area for fishing. So she may be a worker there or a fellow like-minded, sinking-addicted gal. I hope for the latter!

The AI can come up with some very interesting designs for clothing, even if the are impractical. There were better, more realistic, outfits, but a lot of them weren't postable due to things like extraneous limbs etc, so I settled for nice pictures. For the thumbnail you attached, if you think of the clumps in the background as lumps of seaweed, this looks very much like an estuary mud flat I know of, so to me it is realistic and quite suitable for fishing. In fact I was reading an article a while back about a fisherman who got stuck in the mud there and had to be rescued. My first thought was "wimp, too fussy to sit down and pull his legs out!". I've been through that bit myself and it's not that bad.

Watch this last one, if she invites you round for a meal and there are fava beans and a nice Chianti run like crazy!

Yikes, the crazy eyes..! The teeth..! The...the...chest thingies! Are they outside her catsuit, or are they bursting forth from within???

Did the AI do it like this on its own or did you prompt it that way?
If it's the former, then maybe it's time to be afraid.
Still lovely to look at though!

Well, I nicknamed her "Fangs", she is mostly a result of the AI (pushed fairly firmly by my prompt, I was using a lot of weighting terms). The "adornments" on her chest, sort of, come from the prompt. SD tends to smooth over nipples under clothing even with thin material, so I usually add a "nipples" to the end of my prompt to prevent this. I tend to down-weight the term e.g. "(nipples:0.9)" or so depending on results, but SD, sometimes, applies its own interpretation and adds "features" on the outside of the clothing, e.g. valves, buttons, seams etc. The teeth and eyes though, all the AI, so, ] yes be afraid.

[cerberus]

Seems everybody but me has been having a go at Vera Miles pictures. So, I've bowed to the inevitable and done a few myself. I started doing fairly normal ones (the third) without mentioning Tarzan so got a more mature Vera. After specifying Tarzan, I got a younger Vera, then I had a play with colour pop (but quite subtle) and muted colours (see last one). I was quite pleased with the outcomes.

00105-3688602789.png

00109-4060325525.png

00803-958910545.png

00897-3113643844.png

01231-3000042687.png

[mudmaiden]

These are Great!