Futa Muscle Mom - More Vids 2

Edit: if you downloaded this within 45 minutes of posting, I'm sorry, a jarring cut got inserted into the loop-splashtitfuck_2025-05-30_121915_UP_00001_combined.mp4 video, I just fixed it and replaced the faulty video. So, if you downloaded within 45 minutes of the original posting, please redownload to get the fixed version. Thanks!

Horny TL;DR: what do futa muscle mom's do at night? Goon hard.

Content Delivered To:

• Man Balls Archive

  • 1 video added

• Mom Balls Archive

  • 28 videos (10-35 seconds long) added to: Releases\2025\008-Muscle Futa Mom\Video\Combined

• Teacher Balls Archive

  • See Projects\2025\GA-GothMuscleMom

    • Full sync of the project folder as it currently stands (2,373 files added, images and videos), so a massive amount of additional content, much of it in an unfinished state, but also the source files for everything I've released for this project so far. Go nuts poking through it and saving what you want, it is probably overwhelming (it is for me)!

I'm overwhelmed by how much content I have produced, the hard part, again, is sorting through it and post processing the clips to a finished state. Sure, I could just pop out little 5 second clips endlessly, and that would be fine, there are plenty of AI creators who do that, but I'm really interested in making little longer scenes. The end result is that I'm ending up with a clog at the bottleneck of me sorting through and releasing stuff. What I probably need to do is slow down generating new content and focus more on the work of getting what I have out the door. Today, I did a marathon of work to get this batch ready, so I'll probably take it easy this weekend.

I went through the Futa Muscle Mom project folder and did a bunch of cleanup and decided which clips were ready for prime time and which ones I'd like to work on a bit more. I ended up with 28 I felt were ready for release. So, sorry for sitting on some of these for so long. I realize that my technique has gotten better since then, but these were still pretty good. These generations were when I was really pushing the limits and see how much drift would be acceptable. As a result, you'll have some 20-30 second clips that kinda go off the rails before the end with some weirdness: mainly with the character's appearance changing and with penis tips losing their definition. I cut the clips off before they went too far, in some cases I'd go for another 2-6 five sec generations seeing how far I could take it. Now with a Start+End Frame workflow, I have a potential "check" on drift, allowing me to organically return to a source image or earlier part of the sequence to restore appearance before it goes too far. I selected another 5 clips I wanted to do this with that I'm not releasing today. And I'll probably go back and redo "spray" and "thickcum" and "undrcover" as well since the drift was pretty bad in those.

But, I am releasing my first example of use of Start+End Frame with the "loop-splashtitfuck_2025-05-30_121915_UP_00001_combined.mp4" to make a 35 second looping video. This was my first experiment, I liked the titfuck, but I couldn't get her to cum in a satisfying way, but the initial frame had a nice cum, so I was like, "can't I go back to that?" That's when I decided to investigate doing a Wan 2.1 Start+End Frame workflow. Support for End frame stuff isn't great, it requires custom nodes, and the workflow I currently use is highly customized meaning I'd have to figure out a way to either incorporate it manually or customize one that features end frame to my liking. I decided to opt for the former since it seemed simpler. Problem is, the VRAM requirements went up dramatically, so I've had to run it in the Cloud so far on an RTX 4060 with 24 GB of VRAM, it gets pricy, but since I'm only doing it for some generations, it isn't too bad.

Other than that, you'll notice some "fogging" when transitioning between generations at times. I believe this is an issue with Wan's denoising process and it just crops up randomly no matter what I do it seems. It seems to depend mostly on the source image and Wan really does it a lot to the source images of the muscle mom. Frustrating, but I don't think it detracts too much.

In any case, hope you guys enjoy! In other news I decided to see if I could finally get POV docking to work, especially since there is no LoRA for this concept, well, with a missionary LoRA and a bit of prompting magic, I'm getting good results that I hope to share soon. I'm also refining my techniques getting the Mom Balls Sex videos looking good. All the while, patiently waiting for the sound design on my long scene...