Today is the day I am going to introduce you to one of Newton's laws: Conservation of Energy. Lucky for you, I didn't bring my Newton’s Cradle with me, but assistant Ashe and Pharah agreed to a little... demonstration. The potential energy of my hips turns into kinetic energy as I swing, passing it down the line as I go balls deep.

Professor's Widow lesson

Hi, how are you guys doing, how’s your summer going?

I finally managed to assemble all of the pieces, and let me tell you - I wasn't ready for it. I was dodging Jensen's advanced course for years, but this project actually made me pay him and buy a new GPU (I'll tell you more about that later).

I actually forgot how to properly export a video from Resolve so that it maintains the right colors and all the data needed for transcoding into different formats. Because of that, I spent the last four hours just sitting there trying different settings. (It's now more than that, but it's done)

A lot has happened during the project, and I'd like to briefly talk about some things. Actually, let me just yapp about the overall timeline and see where it goes.

1. Animation Evolution

If you remember, at the beginning it was a single shot: four characters on screen for 10 seconds.

Now that I see that, I wanted to make a line of 4 but it didn't look good, so I had to reposition Domina on the floor

That's it, that was suppose to be whole piece.I thought the overall idea was great, so I added an intro, then some more, then another shot, then one more, and... At that time, it was about 900 frames (30s), and I was more or less satisfied with the way it looked.

That was about a month ago.

Time to render this out and... 15–18 minutes per frame. "Hmm," I thought, "Okay, that's not good, but it's not terrible, right?"

(Well, actually -_-! ) It IS terrible. I did the math, and there was NO WAY I was going to finish this on my GPU. I was thinking to myself, "Did Jensen finally get me?"

And he did.

The following week, I was basically looking into different GPUs, monitoring prices, and decided to go with the 5070 Ti. It could be better, it could be worse, but for what I do, it’ll be more than enough. Render times went from 15–18 minutes down to 3–4 minutes per frame, and everything can now fit inside the VRAM.

2. Universal Rig

During animation, it became obvious to me that something had to be done regarding this. Think about it: you are animating four characters, each slightly different from the others in terms of controls, materials, and capabilities. If you want to add something, you potentially need to do it four times. Oh my god, this is such a waste of time. It’s a dead end. Also, you never know if the thing you want to do will be doable with a given rig without additional modifications, which puts you in a weird spot. And all of this is understandable - you are getting models as is, it's now up to you to debug and modify them as you see fit.

I did some digging, and I think I more or less know what I want from the rig and how to go about it. At first, I was adamant about creating a custom topology and deformation rig, but I've changed my mind.

I am not sure what I will pick yet, but what I do know is that it won't be DAZ. I don't like their approach of using corrective shapekeys to fix deformations instead of corrective bones. You balloon the .blend file and remove the easy way to attach cloth to the model via weight transfer - and for what? Nah.

Metahuman looks interesting, but it's so sloooooow. Also, I feel like the more you look into how it's done, the more complicated it becomes.

Then there is the Character Creator rig/model, but I am not sure I want to deal with them.

3. Compositing and Editing

I wanted to talk a bit about color science, but I just can't do it. I hate everything about it, and the post just becomes too toxic for no reason.

I wanted to show some of the bugs but there were so many of them, let's just skip straight to color grading: it was actually quite interesting this time. As you know, you often don't get perfect and consistent lighting straight from the render. Some shots were overlit; some had different lighting or colors. During this step, I was trying to bring them together to achieve a persistent color, look, and mood between each shot.

I tried it, and everything looked even worse than before. I went with saturated orange palette to make it more cartoony, but it just wasn't looking good. I was trying to salvage it, and I accidentally changed something, getting a completely different look - much darker and brown instead of orange.

Video available on the cloud drive or in the attachments down below

I knew that was it the moment I saw it.

I redid everything, and this is what you see in the final version. The only shot that was an L is the one with the blue screen on.

What else? Oh, take a look at this, see if anything is wrong.

Look at this string that clips inside the nipple. What is this?! And it's right in your face. Now I need to go into Blender to fix it, wasting so much time. Well, not exactly. If it's not completely broken, you can fix it inside the compositor; you just need to go frame by frame and directly paint what should be there. Luckily for me, it was a very short clipping, so I just went and painted over the footage, saving some time.

What you can't fix easily is this; took me quite some time to figure out how to make this and to come up with a fitting material.

Audio - nothing special. You just sit there alone, trying to come up with something interesting to fill the void. That's pretty much it. What I don't get is how people can charge like $15 for a 30-second video? It's nothing. I'm curious what kind of quality you would get for that, but I think I already know the answer.

Btw, if you can I'd like to hear your input about audio levels. Is it good, bad? I've tested it on a couple devices, it should be alright.

4. Commissions

I'll have to do them; I need to offset my losses.

Right now, I am thinking about the format and the amount I can safely take. There will probably be a separate pledge with a single slot, so if you are an oil prince, you can just buy that.

I am not counting on it being taken, so there should be another way to pick winners. Since everybody loves gambling, I am thinking of a wheel - you spin it, you get a winner. This is just the general vibe; now I just need to think about the details like: how many slots, both tiers or not, for how long, and so on.

What's next

This project looked great on paper, but the actual realization was such a drag.

I am so glad it's finally over so I can move on to something else. I’m thinking about making one with Venom and another with the Witcher girls. The plan should become a bit clearer once I sort out the commissions and the "wheel thing," but for now, this is the general direction.