Progress Report - December 24, 2025

Hello everyone!

I hope you're all having a great end of the year, and enjoying time with your friends, families, and loved ones. I come bearing Christmas gifts in the form of a Progress Report! Whether or not that's equivalent to a lump of coal, I leave as an exercise for the reader.

This past week was mostly spent on more Blender simulations and game brainstorming. Of the handful of tests I did, far and away the most successful was the hair-sim tests, seen here.

I can't overstate how excited I was to see the hair-flip just work. There is zero manual input on my end, everything you see here is driven 100% by the cloth simulation, physics interactions, and the animation on Anna's head. Both the torus pulling her hair over her shoulder, and her just flipping her head to bring her hair over, is 100% physically simulated.

I don't even want to imagine how painful it'd be for me to animate that by hand, and have it come out looking even half as decent. The nuances of her hair bundling up as the torus picks it up, or the hair bouncing off her shoulders as it comes down? Agony to imagine.

Hair simulation was the original impetus for my Blender experiments in the first place. It took me three days to manually animate the hair in Pharmmetra, and do my best to roughly approximate how the hair would fall around their shoulders within clipping into it all. And the entire time, I was just thinking to myself, "this would be so much easier if I could just simulate this."

I learned a few interesting things while doing this that I think I can apply to other experiments as well. The major takeaways I got is that cloth simulation in Blender just seems to overall behave better than soft-body simulation. I originally tried to do this with soft-body, since that makes sense to me, but it ran into the same issue my soft-body breast tests had - the hair kept wanting to collapse in on itself, which isn't exactly desired.

The muscle meshes driving the hair are simple formed cylinders. The simpler geometry allows the simulation to run quickly and effectively.

By switching over to a cloth simulation though, I was able to apply air pressure to the muscle volume, which encouraged it to keep its shape under gravity, only deforming gently as physics collided with it - which is how hair (and breasts) should behave.

As a bonus, cloth simulations are magnitudes faster to calculate. I was able to test and tune this at almost-real-time, whereas with soft-body it was very much a "set up parameters, hit bake, go get a drink and watch something on YoUTube while it processes, then check after a few minutes."

I also learned that while Surface Deform is much quicker to set up, and is much more forgiving, it does not behave nearly as well as Mesh Deform, when it comes to translating muscle animation to a complex mesh like Anna's braids. Her hair had a tendency to just... explode... under Surface Deform.

Mesh Deform works much better, as you can see in the opening GIF. However, I quickly learned it is extremely picky. Any vertex that isn't entirely encapsulated by the muscle mesh simply will not bind, causing the hair to explode in an unsavory way.

The single hair vertex, colored in orange here, is barely poking outside the muscle mesh, colored in black for contrast. Yet that is enough for the vertex to not bind, causing the hair to explode.

Unfortunately it seems like Blender has no way to automatically detect any such clipping, meaning I have to resort to using my eyes and trial-and-error. Luckily, unbound vertices are easy to find due to the aforementioned explosion, and Blender can be configured to allow you to select vertices mid-simulation. So I'd bind, run the simulation, wait for the hair to explode, select the problem vertices, cancel the simulation, adjust the muscle to fit the problem vertices, and try again.

Not an ideal solution, but a simple one at least. Doesn't take much in the way of skill, just patience. I will definitely be looking into ways to automate this in the future.

I want to take what I learned about cloth simulations and the Mesh Deform modifier here and attempt the breast-jiggle test again. The ability to fill the cloth muscle with air pressure to avoid it from collapsing in on itself, I think, will be the key: provide with just enough air pressure to stop the breasts from collapsing in on themselves, but give the cloth muscle enough elasticity such that it will easily cave and squeeze when physics objects collide with them - be that pressing breasts against a window-pane, or a greedy gripper grabbing a handful of tiddy.

However, the next Blender test will be exploring spring constraints. The cloth simulation works fantastically for hair that's supposed to fall straight down, like ponytails. But for hair that is supposed to retain its shape while it reacts to gravity, like bangs, it doesn't work so well - it gives a very "wet hair" effect, where the bangs just droop straight down and press up against the skin. Works great for scenes where characters come out of a shower, but not much else.

In Source, this is achieved through spring bones. Under no external forces and assuming a default of gravity at the character's feet (EG standing up, as opposed to hanging upside-down or laying on their side, for example), the spring bones maintain their default position. It is only as external forces deviate from that standard assumption that the spring bones react.

I want to explore if Blender has a similar mechanism. It works fantastically in Source, but the problem of course is the lack of collisions in SFM. If I can get spring constraints in Blender that also have collisions, then that'd be golden. No more having to manually pull bangs out of character's eyeballs. What a world that'd be.

Beyond Blender simulations, I've still been slowly figuring out the lewd UE5 game design, for bringing the Saturday streams back in 2025.

Amusingly, the image you see above was originally made as part of a "Should we do X or Y", contrasting a more abstract brothel-sim game and a more focused single-character game. I think I mentioned this to some capacity last week.

The core dilemma in that discussion is that, in the brothel game, due to its natural of managing multiple ladies at once, you can't really get intimately involved in any one activity - there are too many other things for you to pay attention to. By contrast, with the focused game, you can get intimately involved, because you don't have anything else to do but be involved.

As such, I had put together a general concept for the single-character game of all the different activities you can do being mildly involved minigames. As an example, the idea I had for the sex minigame is an extension of the old classic Zone Flash games - you have buttons for a few different positions at a few different speeds, as well as for choosing whether the lady tops or bottoms (whether she sets the speed and pace, or her partner does). There are then two different meters you manage, one being how close your partner is to popping, the other being how much he's enjoying the current activity. The former meter builds up faster with the faster activities, with the slowest activities allowing it to decay; the latter meter starts out high whenever you move to a new position, but decays the longer you stay in a position. The goal of the minigame, then, is to edge out both meters for as long as possible. The higher each meter is, the more points are provided for every second until the partner pops or they lose interest. The more points you generate, the better the tip paid.

Other activities, such as pole-dancing and filming porn, would have similar minigames.

I rather like the high-level idea, and was ready to move forward with it. But one thing kept snagging my mind, which was character customization. A big part of my plans for the brothel simulation angle is that your ladies have a slew of different mutable physical traits - hair color, makeup, piercings and tattoos, things like that. Prospective clients would have preferences for different physical traits, and matching clients' preferences up with ladies that meet those preferences would result in some bonuses. Another layer of management for the player to engage with.

But with a single-character focus, all of that is rather moot. Constantly swapping around your one character to match preferences is a bit dull, compared to the brothel sim's having a wide range of ladies and then mixing and matching them.

And so, I set out making the picture for this set asking which way to go.

But as I was putting the image together, the idea occurred to me: my original assumption, that having involved minigames doesn't work in the brothel sim, is erroneous. I think it can work - and the challenge of having to manage multiple minigames at once is a natural difficulty curve. When you have just one lady, it's pretty easy to focus on just one minigame. With two ladies, now you have to split your focus, but it isn't too hard. With 8 ladies though? Well, at this point you're pretty deep into the game, and so hopefully you've built up the skills to be able to manage them.

I think it's a natural progression and difficulty curve. The big key to it all will be timing. If each minigame only takes place across 15 seconds, then it isn't viable to manage 8 of them at once. But if they take place over 2 or 3 minutes, and you can set them up to run for a bit and jump between others before needing to adjust them? It can work. Throw in different activities having different levels of player involvement, and I think it actually works quite fittingly.

So, unless I hear or think of arguments against it, I think that is the path I want to take: brothel sim, with involved but slow minigames.

That's all for now. I will see you all next week, with the last Progress Report for the year! After that, we start January off strong by me resuming work on Overbreed. More on that next week!