Technical Breakdown. Making of a pussy + stuff (Patreon)

Good day.Finally, it’s time to break some things down. I haven’t done this in a while, so there are quite a few things I want to talk about.A few takeaways from the previous animations before moving on to the main thing.

Overwatch Halloween animation

VolumetricsFuck them. It’s not worth it at all. Yes, light shafts and light scattering look beautiful, but for me the performance hit often outweighs the benefits. You can add light shafts later in post, or even fake them in 3D, and it won’t double or triple your render time.If you decide to go with true volumetrics, don’t set the sample count too high. In the end, it will still be a bit grainy no matter how high you go. Render them as a separate layer and blur it to remove the grain.

The grain is removed, but now your edges are blurred as well. Doesn't matter. It's not a mask and when you add everything back together you won't notice anything. What you will notice is grain if you don't somehow handle it. 100%.

First part - blowjob

Basically a stepping stone.You would think it would be an easy thing to do, granting you use a nsfw rig made specifically for that.Here’s the problem - there was no built-in functionality for the lips to stick to a surface, and you can tell it wasn’t built with that in mind.

Now that we’ve established the problem, it needs to be fixed.I had a couple of options to do this:

Shrinkwrap the lips’ bones to a shaft.The cock would move in and out, and the lips would stick to it.Potentially the best method for a mouth?

Use a proxy mesh for the mouth and let it be deformed by the shaft.The tech is similar to the previous one, but the deformer is controlled by cloth simulation instead of bones.

Manual animation.

I tried the first two approaches, but I couldn’t make them work at the time. Most likely skill issues - plus I wasn’t sure how to add this functionality to an already made rig without accidentally breaking it.

That left me with the manual approach. To make things easier, I created four control bones - two to control the top and bottom lips, and two to adjust the corners if needed. Then it’s just a matter of time: go frame by frame and adjust them to make it feel like they’re always on the shaft.

Basically, it doesn’t really matter how it’s done. The point was to decrease the number of bones you need to keyframe. Other than that, you can do whatever you want - you could even handle it with a single bone if you really wanted to.

Now, why is this important?Because a pussy is basically the same thing, alright? It’s an opening that needs to deform around an external object.

Second part - pussy time

For the second part, I needed a working pussy. The problem was exactly the same as with the face - the default rig wasn’t enough to do it the way I wanted. It was a standard shape-key-based rig with open/close states and an option to translate it. You can’t really sell motion with that.

The options were the same as for the lips.

I knew people often use a donut shape as a proxy deformer, but I didn’t like it after trying it. It was too shallow and often got turned inside out. While you can deal with it, there’s really no point.

Then there was a tube. It was much more stable, but still, something felt off. There was no point in having it extend all the way down if the only part that actually needs to move is the entrance. You’re just wasting computation time, and the deeper part of the mesh ends up affecting the entrance.

Important: while testing, use a realistic shape for the thing that’s easing in. If it’s blunt, make it blunt. If it’s pointy, make it pointy. Otherwise, your testing might be skewed and, potentially, it won’t work at all with a different shape.This cylinder is too blunt and doesn’t look like a shaft - it was just tearing the mesh apart when trying to ease in.

Now, the donut is too shallow, the tube is too deep, and the truth is somewhere in between.Here is what I've got:

Not shallow enough to turn inside out, and deep enough to capture most of the entrance deformation. It could have worked as is, but you can retopologize it - reduce the poly count and create proper edge loops.

Perfect.Now it’s pretty much the same process as with any other proxy mesh:

Data transfer weights from main mesh to the proxy.

Bind the main mesh to the proxy.

Important: to get the best transfer and binding results, do the following first:

Select your rig, open up the pussy, about 50%, and open up the legs, like 60-90 degrees.

Create a new shape key for your proxy pussy mesh and make it overlay/fit your main mesh.

Basically, you want to split the main pussy into left and right parts before doing any transferring or binding. This avoids incorrect weight transfers when the mesh is closed and vertices are too close to each other. You want to avoid situations where the left side of the proxy deforms the right side of the main mesh, or vice versa.

Something like this:

Now it’s up to you to correct weight pairing where needed and run the simulations. Use smooth corrective modifiers where necessary, and it should be fine.

Put it all together and you end up with something like this:

How about that? Not so bad for someone who has nothing to do with rigging.This took eternity, see you later.

o7