A Pandoras Box of Very NSFW short videos and other random clips.

[BlackOut]

So, in an effort to share some of what I've been working on and the findings I've made along the way I decided to open up a new topic.

I've been using a mix of Krea and Luma - although I am still keeping an open mind and actively looking around to see what other options are out in there in the AI market. While Krea is a pretty solid tool for working on animations, it does take some time to get the result you want. Also, one of the bigger drawbacks with Krea is it does not seem to want to maintain the visual integrity of the images that you use as keyframes. Fingers crossed that it's an issue they resolve over time, but for now Luma is better at keeping the consistency of the images you use.

Now onto some examples.

This first one gets a special shout out to jlocke who so kindly dug up the old archive of this particular set of images featuring Kelsie Chambers as the Scarlet Avenger. It took quite a bit of time, effort, and editing of the images to - trick - Luma into animating them since it kept flagging them as inappropriate. (Something that Krea does not seem to do)

Once I managed to sneak a few in, I took the animations over to Krea and enhanced them.

So without further ado, here's a blast from the past brought to life.

[Mr. X]

Kelsie Chambers! Great work!

[BlackOut]

Imgur seems to be giving me issues so until I find a better option, here's a link to something created via Krea from a single illustration.

https://imgchest.com/p/md7oj5zqdyp

[BlackOut]

Kelsie Chambers! Great work!

She's always been one of my absolute favorites in the genre.

https://imgchest.com/p/5xy2lnzm94l

https://imgchest.com/p/agyvejzbb48

[infogeek247]

Very nice, thank you. What keywords did you use?

[BlackOut]

If I'm using a picture reference the key words don't seem to be an integral factor in what the AI often produces. Usually I try to produce images and use them as references for an animation. The thing that's nice about Krea is that it has a real-time rendering that adjusts directly before your eyes with every keystroke you enter. In essence Krea allows you to watch the computer paint.

I'll probably stick with Krea but I've found Bings AI generator to be pretty solid (despite being very picky about prompts that might hint towards anything remotely lewd)

[BlackOut]

The more I work with Krea, the more impressed I am getting. This latest tidbit was made entirely on Krea's platform

slime time.mp4

(2.18 MiB) Downloaded 684 times

[BlackOut]

This one took some tinkering in the animation department, video kept wanting to render ghosts. Finally, though I managed to adjust things and came out with this. I like to think of it as Wonder Woman's fall to the Son of Ares...

[BlackOut]

Been working with an AI renderer called Perchance - It's a fairly decent free platform that allows for NSFW content and you don't even have to sign up to use it. Tossed a few images I made into Luma and came out with a handful of decent animations of Wonder Woman sinking into quicksand.

[Ernie]

Other than the first one, these are all kinda creepy, uncanny valley-ish animations.

[jlocke]

This one took some tinkering in the animation department, video kept wanting to render ghosts. Finally, though I managed to adjust things and came out with this. I like to think of it as Wonder Woman's fall to the Son of Ares...

Love this! Lots of potential uses. Which program did you use to generate original AI image, or was this entirely in a video program?

[BlackOut]

This one took some tinkering in the animation department, video kept wanting to render ghosts. Finally, though I managed to adjust things and came out with this. I like to think of it as Wonder Woman's fall to the Son of Ares...

Love this! Lots of potential uses. Which program did you use to generate original AI image, or was this entirely in a video program?

This one was done completely in Krea. I made the image based off another piece of similar art and then for the animation I just used the same piece as a key frame multiple times. The keyframe prompt also gives a small degree of additional control, like making the man shirtless.

Other than the first one, these are all kinda creepy, uncanny valley-ish animations.

Most of them were done via Luma, it has a habit of slowing things down. Krea does as well, but it's a little bit more controllable. Still, ultimately the AI kind of decides what's going to happen.

[BlackOut]

Next entry, put together this evening via Krea - I used a prompt and a few shapes to get the initial keyframes for the sequence, added a few more prompts and played around with the settings and swapped a few frames until...

Not everyone believed the whispers about the existence of a secretive German cult that was devoted to unearthing the darkest of mystical secrets. In an alternate reality, Wonder Woman would never be able to share the truth of their existence, not after they imprisoned her in the shadow realm.

video-enhanced(8).mp4

(15.69 MiB) Downloaded 621 times

[BlackOut]

One of the major drawbacks for AI that I've encountered is having a model that looks consistent - I do believe I've discovered a technique in Krea that has resolved that issue.

EroticRender01.mp4

(14.62 MiB) Downloaded 649 times

[BlackOut]

Another render done earlier today via Krea. A little belly dance and sway of the hips.

EroticWWBellyDance.gif (23.21 MiB) Viewed 44016 times

[sneakly]

considering how bad it was a year ago, this is developing pretty quick.

[BlackOut]

considering how bad it was a year ago, this is developing pretty quick.

You can say that again...

EroticSideView.gif (20.25 MiB) Viewed 43991 times

[BlackOut]

A preview image for a future vacuum bag scene I'm keen to render.

80d1957c-5ff3-458f-8e0a-2a4130bc1e3e.png (712.72 KiB) Viewed 43973 times

[BlackOut]

Better than some of the other update costumes that I've seen Wonder Woman wearing. Or maybe I just have a thing for latex.

ezgif-4-4751ece2f4.gif (14.14 MiB) Viewed 43863 times

[jlocke]

Better than some of the other update costumes that I've seen Wonder Woman wearing. Or maybe I just have a thing for latex.

ezgif-4-4751ece2f4.gif

I like this one too! I was messing with Krea.AI (free) and I can't get the video program to stop changing my model. It morphs Lynda Carter into a different face with smaller breasts each time. Any advice for keeping the model exactly the same as the uploaded image?

[BlackOut]

Better than some of the other update costumes that I've seen Wonder Woman wearing. Or maybe I just have a thing for latex.

ezgif-4-4751ece2f4.gif

I like this one too! I was messing with Krea.AI (free) and I can't get the video program to stop changing my model. It morphs Lynda Carter into a different face with smaller breasts each time. Any advice for keeping the model exactly the same as the uploaded image?

So it's a good deal of trial and error, and since I'm signed up for the Pro Plan I have unlimited renders, animations, and enhancements at my disposal. I've also seen that allows me to train the AI to recognize certain aesthetics, which I plan to delve into sometime in the near future. There's even a seed option that can be used in the legacy apps section which should make it easier to keep a model consistent. Honestly, it's all new to me, so I'm just kind of learning along the way.

But, this is one trick I've learned over the time that I've been using Krea. Under the Realtime Editor, there is an option for Image to Image creation. I typically start with the prompt, editing it along the way until I start to get a result that meshes with what I'm after. Then it's just a matter of refreshing the variation button until you get a result that you find to your liking. That's step one.

Screenshot 2024-07-20 at 22-54-04 KREA AI - Realtime Editor.png (901.7 KiB) Viewed 43819 times

Step two, I start to draw on the blank page to the left of the image. You can use it to create a degree of motion, to get colors right in certain parts of the image, and more. But once I have that first image, I do a few strokes on that side and get the image to shift a little while the model remains fairly consistent. Sometimes the AI changes things drastically, but there's unlimited amounts of undoes at your disposal so you can always move back. I save what I like, clean the canvas, do it again, until I have a handful of results that I feel I can animate with.

Screenshot 2024-07-20 at 23-04-36 KREA AI - Realtime Editor.png (1.01 MiB) Viewed 43819 times

Finally, once you are in the animation stage itself, the sliders (from what I understand) relate to the prompts. Each keyframe should have a prompt beneath and the higher the sliders value, the more the prompt is in effect. I find it helpful to really hammer home things that are important, by typing them multiple times, so if you don't want Lynda Carter to morph, I would try typing that maybe 5 or 6 times just so the AI hopefully gets the point. Also, it's a good idea to keep the Motion Intensity in the 40 to 60 range, anything higher and you can start to get a maelstrom kind of effect that takes over the animation.

Lastly, I do spend quite a bit of time on the animations, like I said, I have unlimited amounts of renders I can do so I keep plugging away and making little adjustments, trying different images, etc, until I have a result I'm content with.

[BlackOut]

Just a quick test based off of some of the prior batch of rendered images I've been tinkering around with. Motion Intensity set to sixty, and the video is looped, the prompt is...

Several large snakes are coiled tightly around Wonder Woman, she is rolling from side to side on the floor, she is struggling against several snakes that are wrapped tightly around her body several times and squeezing her. She has a desperate expression on her face and has a helpless body language. The background is black. The snakes are wrapped around her body multiple times. The snakes are squeezing tight around her body, she is helpless, the snakes are around her waist and slithering all over her body. She is laying down on her back against the floor, writhing and struggling as multiple snakes defeat her. She is gasping for air, she is gasping for air, the background is black, the background is black, She is on her back and wrapped up tightly in the grasp of several snakes, the snakes are constricting tightly around her body, she is twisting and turning about, the snakes are firmly coiled tight around her body. Slow motion animation, very slow motion animation, no morphing, objects remain consistent and do not transform or morph.

snakes01.gif (19.17 MiB) Viewed 43816 times

[jlocke]

Better than some of the other update costumes that I've seen Wonder Woman wearing. Or maybe I just have a thing for latex.

ezgif-4-4751ece2f4.gif

I like this one too! I was messing with Krea.AI (free) and I can't get the video program to stop changing my model. It morphs Lynda Carter into a different face with smaller breasts each time. Any advice for keeping the model exactly the same as the uploaded image?

So it's a good deal of trial and error, and since I'm signed up for the Pro Plan I have unlimited renders, animations, and enhancements at my disposal. I've also seen that allows me to train the AI to recognize certain aesthetics, which I plan to delve into sometime in the near future. There's even a seed option that can be used in the legacy apps section which should make it easier to keep a model consistent. Honestly, it's all new to me, so I'm just kind of learning along the way.

But, this is one trick I've learned over the time that I've been using Krea. Under the Realtime Editor, there is an option for Image to Image creation. I typically start with the prompt, editing it along the way until I start to get a result that meshes with what I'm after. Then it's just a matter of refreshing the variation button until you get a result that you find to your liking. That's step one.

Screenshot 2024-07-20 at 22-54-04 KREA AI - Realtime Editor.png

Step two, I start to draw on the blank page to the left of the image. You can use it to create a degree of motion, to get colors right in certain parts of the image, and more. But once I have that first image, I do a few strokes on that side and get the image to shift a little while the model remains fairly consistent. Sometimes the AI changes things drastically, but there's unlimited amounts of undoes at your disposal so you can always move back. I save what I like, clean the canvas, do it again, until I have a handful of results that I feel I can animate with.

Screenshot 2024-07-20 at 23-04-36 KREA AI - Realtime Editor.png

Finally, once you are in the animation stage itself, the sliders (from what I understand) relate to the prompts. Each keyframe should have a prompt beneath and the higher the sliders value, the more the prompt is in effect. I find it helpful to really hammer home things that are important, by typing them multiple times, so if you don't want Lynda Carter to morph, I would try typing that maybe 5 or 6 times just so the AI hopefully gets the point. Also, it's a good idea to keep the Motion Intensity in the 40 to 60 range, anything higher and you can start to get a maelstrom kind of effect that takes over the animation.

Lastly, I do spend quite a bit of time on the animations, like I said, I have unlimited amounts of renders I can do so I keep plugging away and making little adjustments, trying different images, etc, until I have a result I'm content with.

Thank you! That's very helpful. I think my best bet since I'm on the free Krea is to stick with some of my free programs so I can refine the instructions each time. Keep up the great work though! Love the Wonder Woman in Peril stuff!

[BlackOut]

Often times an animation is close, but portions aren't right, so I'll drop it into an editor and crop out the bad stuff, maybe add a little disco effect.

[BlackOut]

Every now and then I'll do something on my phone, which currently severely limits my ability to add adjustments and prompts to the project. Still, sometimes the AI does something nice looking.

[BlackOut]

[BlackOut]

ezgif-2-46263062b9.gif (18.59 MiB) Viewed 43640 times

[jlocke]

ezgif-2-46263062b9.gif

Very cool! Love this one!

[infogeek247]

Thank you for all the information.

[BlackOut]

Here's one that turned out pretty well, although I can't take credit for the image itself.

ezgif-2-7b50021806.gif (19.99 MiB) Viewed 43413 times

[BlackOut]

Another one that turned out well, again no credit taken for the initial image

ezgif-6-9fec6afbe3.gif (11.4 MiB) Viewed 43405 times

[sneakly]

Some of these are really great. It will be interesting to see what can be done in a year or two. right now simple movements seem to be the limits. Eventually there will be the ability to render an animated story board and ultimately an animated story. Good work

[BlackOut]

Just been playing around today, no prompts, just letting the AI do it's thing to some random pictures I toss its way

ezgif-7-32127aba05.gif (22 MiB) Viewed 43351 times

[BlackOut]

I should start a game called "Guess the picture" No prizes, unless you count an ego boost as a prize...

ezgif-1-15f036d08e.gif (11.34 MiB) Viewed 43340 times

[BlackOut]

Quite the stunning pose if I do say

ezgif-7-a9898a38fd.gif (20.61 MiB) Viewed 43333 times

[infogeek247]

What product did you use to get the mobile and moving lasso please?

What aI?

[BlackOut]

What product did you use to get the mobile and moving lasso please?

What aI?

Krea, which I've been using for pretty much all but a few of my animations.

[BlackOut]

Started working on a series of these, an homage to "The Boy Who Knew Her Secret"

[BlackOut]

Another addition to the "Boy Who Knew Her Secret" set.

ezgif-7-6e30e2dfcd.gif (10.8 MiB) Viewed 43105 times

[BlackOut]

Just a random hypno scene that I rather liked.

ezgif-5-c733e49434.gif (1.28 MiB) Viewed 43100 times

[BlackOut]

These aren't quite perfect, but for the amount of effort (almost none) Krea did a fine job rendering this out.

downloadedImage (41).png (269.97 KiB) Viewed 43034 times

downloadedImage (42).png (268.08 KiB) Viewed 43034 times

downloadedImage (43).png (259.91 KiB) Viewed 43034 times

downloadedImage (44).png (275.17 KiB) Viewed 43034 times

[jlocke]

Another addition to the "Boy Who Knew Her Secret" set.

ezgif-7-6e30e2dfcd.gif

This is my favorite so far. The original episode was full of peril. When Diana has the pyramid in her hand and is trying to shake free and can't, great stuff. Love the pyramid special effects on this one.

[BlackOut]

Another addition to the "Boy Who Knew Her Secret" set.

ezgif-7-6e30e2dfcd.gif

This is my favorite so far. The original episode was full of peril. When Diana has the pyramid in her hand and is trying to shake free and can't, great stuff. Love the pyramid special effects on this one.

I was planning to work on that series more, but I got side tracked... Honestly, surprised that I was able to achieve even this result.

ezgif-4-8cb18d0c76.gif (10.59 MiB) Viewed 42991 times

[tmon]

WOW!!!!

Another addition to the "Boy Who Knew Her Secret" set.

ezgif-7-6e30e2dfcd.gif

[jlocke]

I found one! I think. I'm pretty sure this one is from an old starryai image I did from a photoshopped Erica Campbell.

erica01.png (524.33 KiB) Viewed 42634 times

ezgif-6-9fec6afbe3.gif (11.4 MiB) Viewed 42634 times

Original and photoshop:

e8c2.jpg (193.7 KiB) Viewed 42634 times

e1.jpg (186.78 KiB) Viewed 42634 times

[BlackOut]

I found one! I think. I'm pretty sure this one is from an old starryai image I did from a photoshopped Erica Campbell.

erica01.pngezgif-6-9fec6afbe3.gif


Original and photoshop:

e8c2.jpg

e1.jpg

You are correct! Good eye (and memory), also thanks for providing that image, I simply used it as a key frame multiple times and rendered it on my phone - maybe I'll play around with it again on my actual computer and see if I can get even better results.

Current project in the works is an electrocution chair peril for Wonder Woman, here's an initial image that I'm going to try to use as a base template.

downloadedImage - 2024-07-26T063901.694.png (1.74 MiB) Viewed 42813 times

[BlackOut]

Well, ladies and gentlemen, I have some truly exciting and mind-blowing news to drop here. I am definitely not kidding around, this is one exciting discovery that I have been working towards understanding just how far it can be pushed. Below I am going to share a gif (which was a video file initially - but switched to gif format for sharing purposes here). This gif I did not render images for, this gif was not animated by virtue of keyframes, this gif was created from a video file of a porn star having sex with a man that I dropped into the Krea enhancement tool. Now, the Krea enhancement tool has a prompt section, it usually fills it out with some algorithm that studies the video. This time I changed the prompt and wanted to see if I could add a tiara to her forehead. It did, kind of, but that initial result got me curious. There's sliders involved in the enhancement process, and while I won't go into vivid detail just yet, let's just say by virtue of the prompt I hammered home the point that this was Wonder Woman in the video and tinkered with the slider values for several tests. I'm still playing around with this, but the result here is pretty amazing to know that you can take a video and make such changes.

Wonder Woman Vid Test.gif (19.84 MiB) Viewed 42759 times

[BlackOut]

Another addition done via Krea by virtue of several images I rendered via AI

ezgif-7-68f9695230.gif (22.18 MiB) Viewed 42458 times

[BlackOut]

Remember that part where I mentioned being able to add clothing via prompts to a video? Seems I can also do the reverse.

ezgif-4-92f6278882.gif (3.21 MiB) Viewed 42188 times

[DonShip]

Remember that part where I mentioned being able to add clothing via prompts to a video? Seems I can also do the reverse. ezgif-4-92f6278882.gif

Sweeet