Home Creators Posts Import Register Favorites Logout

Videos

  • bmtbguymilf2026.mp4

Downloads

Content

In the beniningui... I've been spending so many time rendering fixes for DV32 that, after some findings and inspirations, I managed to and plan Countryside City/Milf arc for Giantess Dream seasons 2 and 3 connecting those characters to the actual timeline of Domina's Valley/Cheap Tricks, a gap that's been there for a while already. The old draft stories will become chapters soon and a 3rd season deepening the Wintergreen Garden Family lore into the canon with the presentation of a proper protagonist for a "nightmare scenario" will be very fun to explore.

This extra time invested in DV 32 happened for multiple reasons like continuation errors, animation errors, light bugs, fixes in the scene, multiple "out of memory" breaks and etc. This reminded me that I'm on a complicated position right now: my setup has limitations. Those will be fixed with time. Hopefully. I hope.

Speaking of which, I also tested some online and local AI tools. it feels much better now and is very promising: it evolved a lot and could fix my problems with animations... I mean, I can render a massive scene with more quality and less errors by feeding it with keyframes and filling in with prompts, references and trained models, instead of rendering for 24 hours and finding out 3 out of 20 frames are glitching and screwing the animation.

BUT(T) sure, AI is not perfect, most of the time it breaks the scene and adds crap, but I managed to do simple things that gave me hope. It might be possible to turn those into animated comics and, who knows, one day make Cheap Tricks a soap opera with proper voices and actions?

The potential is there. I just need a 48gb card to run the model I want and invest hundreds of hours training the characters nodes... but that's for later. Testing it while waiting for the queue to clear was fun.

I used an old project where Madison is accompanied with Wendy, that is quite big. In the original idea Maddie was REEEEALY upset with her mother's overdevelopment. It was crazy hard to make it look okay before posting, most of the time something came off, without proper trainings I had to rely on seeds variations because I can't invest big time on adjustments.

I think this is a good example of how things could come out one day.

Enjoy the attachment, it has sound!PS: I'm reviewing the final DV 32 right now. After this I might go for DV 33 and clear out this part of the plot and probably CT 46 then. Unfortunately this knot couldn't be fixed before and this might take some months to fix... but hey, got massive animation knowledge and the Gordon State is free for housing the next NCExtra and future chapters of DV, so I'll call it a win.

Files

Comments

Muad3D

There is potential in AI and I also use it for a good deal of project management and planning, but still feeling very weird about AI images. I just hope whatever you decide to use it for, you won't let it muddle the unique style you have. AI will water down everything it touches (even if it's just interpolating between frames) and it would be a shame to see that happen to your art.

chris

oh wow they are actually going to talk?! How about noises for sex? hot!

Bmtbguy

That's for sure, I don think it classifies as art straight out of a prompt and a reference due to its randomness, BUT I believe that one day, if I'm able to feed my models to it and set voices, moods and style it could become a great tool for bringing my characters to life without me investing 2 weeks for making an animation look good.

Bmtbguy

So, this is a loooooooooooooooooooooong path to walk. First I need the card, test if the model fits in, train in characters, clothes, voices, scenes, moods, variants... it's so much work, time and money to invest in it that this won't happen anytime soon. And then , after all that I'll TRY to think about the X scenes. Just look at the top AI guys, not even them can make it accurately, and I work with massive disparities on a very strict size control, and NONE of the AI works I've witnessed has ever kept even 80% of accuracy. We will do it in the future, the right way, I'll keep studying stuff with the spare time, but unfortunately it will take its time before happening.

Muad3D

With all that effort and investment, it might be more fruitful to animate "by hand" and hire voice actors :D At least then the quality is 100% yours and reproducable, as you created it.

123jb123

for the love of god, dont touch ai. comics are perfect how they are because they are comics. If this turns into ai crap, it will be a no go for so many people, me included.

Bmtbguy

Don't worry, I won't use it unless I can fully control it and make it look and behave exactly the way I want (maybe in more 5 or 10 years lol). I study it because, well, this crap is everywhere and it's been useful for non GTS affairs too... I hated iRay back in the day because it couldn't give good results in the beginning and now, that it evolved, I had to convert everything to it because now it looks good and allows me to render more in less time, but if I learned before I could have invested more in the video card (cheaper in that time) than in the processor. Though cookies...

Pick Me

Ooh I love this for the future! Just imagining the production is awesome