Back to top arrow

Whilst writing more novels is now on the proverbial bookshelf (for now, at least), adapting what has already been written into something visual has been taking up most of my time! I said last blog entry that I was taking a break from making assets for Zuperbuu Works stuff, to focus on some 3D printable things, and for June that is exactly what I did. However, I did still decide to improve on a Zuperbuu Works model! Let’s get into it, shall we?

As you may be aware from the March blog entry, I remade the Jinimaru Base Mesh because the original had notable flaws, a result of it being a Frankenstein monster created from my old Human base Mesh.

The improved Jinimaru model made me greatly consider updating the old Human mesh as well, as both meshes would share the same screen at some point, and having both models on the same quality level was a must. You can read about the Jinimaru model in the old blog here: http://www.zuperbuu.com/uncategorized/areworkandremake/

So how do the new Humans look? Granted the outfits are still old, but the primary focus is on the base body; its topology has been significantly improved, making weight-painting easier, especially the face. The hands and feet have improved proportions, phalanges on the fingers and toes bend appropriately, and whilst it’s not immediately obvious they also now have defined nails.

I tried to mimic the same pose with the new model, but it also has an updated rig, so it was the best I could do! The old model is on the left, new on the right.

With the below image, you can see the drastic improvements to the hands. I tried updating the rig on the old base mesh, but the fingers were too short and stubby and wouldn’t bend appropriately. The finger positions in the OLD model are at their maximum bend before deformities start to happen, whilst the NEW model still has plenty of bendability to go before such deformities happen.

There is a downside to this improved mesh, however; the polycount. Counting ONLY the body (no clothes or hair) the old model was around 9,000 polygons, the new model is a whopping 21,000! Granted, most modern AAA games’ “player” characters are around 5x that number. Still, I will be working to optimise it, especially for background characters or NPCs to use.

Here’s something I don’t show often: the model in motion!

I’m still learning animation, but it shows some of the base mesh’s capabilities. Also, the hair moves now (it doesn’t in these loops, read further to learn why).

I have made bones on the rig that control the hair, so it can bounce, sway, and spring around automatically whilst the character moves, so I don’t have to pose it, or rely on slower physics systems, to make the hair move.
The clothes probably need a remake too, but they can wait. Regardless, I made bones on the rig to control skirts/dresses, and long shirts/jackets/coats that extend beyond the hips. In this character’s case, his shirt has only a teeny-tiny bit of motion on the bottom.

The bones for the hair and skirt are called “wiggle bones” in Blender and work best when not used in looping animations like these, as their positions reset when the loop starts again, so I have disabled them for the above 3 examples.

For game development, I simply turn the wiggle bones off when I export the model’s animations and turn them on again in whatever game engine I choose to use; this way the animation can still loop, but the hair/skirt bones will move independently.

You can see the bones in motion in the below animation. Both the hair and the bottom of the shirt are enabled. Can you see the snapping?

I have a female body made too, I just don’t have any clothing made yet that fits it. The female character, Freija, needs a wardrobe update anyway, so she may be the first to get brand-new clothing meshes!

I did some quick tweaks to my Gakidou character model, to test a few things I learnt with the Human mesh update on the Jinimaru body. Because Gakidou has very elaborate hair, he was the most suitable candidate for these tests.

The first was adding “wiggle” bones to the tail, allowing it to move based on the body’s kinetics. Whilst I like the result, it felt a bit too stiff and needed some tweaking.

The second test was adding improved hair bones. The above animation has them enabled too but they’re not as obvious as the one below. The tail’s stiffness is also a lot more obvious in the running animation.

The final test was with “shell fur” on the body. For those who don’t know, “shell fur” adds detailed fur to a mesh without adding thousands of polygons of physical geometry. The “shell” is made by creating copies of the original mesh, layering it multiple times over the base mesh, with slight spacing between each layer.

Viewing from a distance gives the illusion of 3D fur on the model. When viewing the close-up, you can see the layering (displayed as gaps) that achieves this effect.

This method for creating fur with less geometry is often used in video games, including realistic ones such as Red Dead Redemption 2; all fuzzy animals in the game make use of shell fur!

The shell fur method imports in Unreal Engine 5 and Godot, so I really, REALLY, hope I find one of these engines enjoyable to use. I’m prepared to learn either of them, but it took me 5 years to get comfortable using Blender, and I don’t know if anyone else will have the patience for me to take years to get cosy with a game engine.

Still, should any clients wish for me to make them a model with shell fur, the answer is now a resounding yes! (and it looks WAY better than fur cards!)

I did further testing with the shell fur and wiggle bones and obtained a far more flexible tail as a result, though once I started testing it with running animations it started flying everywhere. So, maybe it’s now a bit too flexible. I still have much to learn.

I had some fun positioning the camera to follow the model, mimicking a game camera. The next step is to get these models into the game engine and see if they even work.

I still cannot answer this question. Right now, I am looking at a replacement game engine to use, as Unity has made repeated poor decisions in the use of their engine; creators having to fork out a fee per installation of their game/project is outrageous, and their claim that it’s what the community wants is ludicrous.
What time and money I spent trying to learn the engine has unfortunately gone to waste, as no two engines are the same. To avoid this setback from occurring again, I am currently browsing my options and testing each engine to see if it feels right for me.

A game developer uses an engine to make their games, but did you know that animators use them too? This is why it is a good idea for me to learn a game engine, as rendering animations in Blender is extremely taxing on the program and my PC, compared to rendering animations in a game engine which is more optimised.

At the moment I am checking out Unreal Engine 5 and Godot, but I am well aware there are other engines out there.

Several models were released over June, including the much-anticipated Version 2 Shenron model; the remake has a LOT of new features that will be seen in all my future 3D printable action-figure works. Most of what I am about to cover has been highlighted in the Patreon page, so I recommend joining up if you’re into my 3D printing work and want to keep updated.

https://www.patreon.com/thomastedwards

Version 2 of Shenron has full ball-and-socket jointing on the arms and legs, the back of the head, and even an articulated jaw. For those who want a little articulation on the limbs, a “light” version is included in the download.
The banana-coloured Shenron, seen in the picture below, uses the “light” limb options and the head with whiskers sculpted onto it.


A few other models that came out were a series of Dinosaur-themed piggybanks; the T.Rex was made free for everyone on Printables/Cults3D/MakerWorld, but the three herbivores are only obtained from Patreon, Cults3D, and Printables for $5.

Presently I have multiple complex models on my plate to finish, but don’t know which to do next. Until my Patreon poll gives me an answer, I cannot say what models will be next.

Sadly I cannot find a more experienced dungeon master to help me proof my module, so I will have to provide it as-is. It’s playable, but the monsters may be too powerful for level 1 characters.

Presently, I cannot say when, or even WHERE, the module will be available.

That wraps up things for the blog so far. It’s my birthday month, so I’d love to spend it getting some progress going in the game-dev or animation department. Luckily you can use game engines to make animations, so if programming a game doesn’t work out, I can still animate!

The content of this site is protected by Copyright.