Feet and Futa

Link to New 17-Minute Video with Still Samples, Updates, Wan2.2.

Added 2025-11-15 02:03:54 +0000 UTC

Where to begin? I managed to accumulate hundreds of Grok Imagine Kissing and Foot-Worship clips that were generated based on my recent DAZ renders. I organized them and put them into a long sequence. The scene starts out with simple kissing between two of my new favorites, Amrita and Selena. As the video moves along, their making out gets more intense and open mouths and touching tongues can be seen. They make it onto the bed and foot worshiping begins. It starts off tame, but Amrita's foot eventually gets soaked with sal1va. The clip finishes off with slow toe-suck1ng.

Grok Imagine really knows how to do kissing scenes and doesn't seem to mind foot worship. It seems to have gotten better over time as if its training on and learning from the images I'm creating and feeding to it.

But what about the NSFW stuff? When are we getting back to X-rated posts? Well, I've got great news. Grok occasionally spits out surprise nudity as you'll see in the upcoming Baldurs Gate video. It also has a hate-on for Witchy Marlene and her very tame foot-tease poses. There's no real way to circumvent the arbitrary censorship other than to use a different platform. I have found that platform.

The holy grail of local video generation created on my computer has been found. I have spent the last couple weeks pouring over Comfy AI and Wan 2.2 video tutorials. I have had 14-hour conversations with Grok about creating the ultimate setup: one that lets me run both of my 5090's at once and takes advantage of my 64GB of VRAM and Terabytes of M2 memory. I have downloaded countless LORA files that will enhance the basic 26 GB safetensor I'm running the system off of. These LORAs range from everything between Foot Worship to Futanari Auto-fell@ti0. The total system is probably over a terabyte of files.

After days of downloading and trial and error, I have a working template and an organized file system. I am ready to rock and roll with this. I've created a few minutes of video so far. Not quite as polished out of the box as Grok Imagine, but close and it's just the snowflake on the tip of the iceberg of what will be possible when I start using these LORAS and get better with the positive and negative prompting. Plus, there's tiddies. I will absolutely be able to create the kind of long-form content that I want, sans the censorship, and relatively quickly. I'm very happy with how fast the NVIDIA cards spit out the videos. Starlight MINI has proven to be the ultimate upscaler and it can polish these files into crisp HD, as it has done for the long video linked here. Topaz Video AI also has an exceptional frame interpolator that can turn 24 FPS into 30, 60, or whatever I want. as you'll see here with basically zero artifacts.

In other news, I have been fully refunded for the overpriced MOCAP system I ordered. Blender just released a billion-parameter motion tracker Addon that can basically replace the need for MOCAP altogether, even when the performer in the reference video is jumping or turning. I can basically make a video of my wife. Trace bones onto her from the starting position and create the exact animations I want based on her movements with no failing or aging equipment to worry about, at home, for free or for a 1-time cost. I love living in the future. Any slight flaws in thew capture can easily be buffed out in post. I then can take the resulting video render and run it through Comfy UI to produce perfect lip sync, hair physics, cloth motion and boob/booty bounce. I can then upscale that to HD or 4K at 30/60FPS. I see a workflow forming and it will empower me to create a frack-ton of content. I still want to greatly improve my knowledge of Blender and Classical Animation, and getting a couple capture suits is still on the table for the next year, but the road to taking what's in my head and bringing it to you has gotten a lot shorter.