Thursday, 8 February 2018


The time has come to finally make what we have envisioned. I told my group that I would like to be in charge of creating and rigging one of the characters. Thus I have been tasked to create the main character.

I started by assessing what needed to be done. The sooner everyone is able to animate the better, this is where I took it upon myself to create a character rig with block shapes and rigid skinning to allow everyone to start animating without waiting on a character model to be finished and textured.

This started with a very simple proxy rig, this was mainly to get pre-vis rolling.

This is following the character designs done by Ashleigh for the son character.

Sticking with pretty much the same proportions throughout development of the character.

As per request I have the squashy pre-vis rig arms as they were important in some of the scenes.
Then it was time to get started on the main proxy rig, the one that will be used to skin the character once he is finished. Excuse the image quality this was a quick mobile upload to keep my group updated on progress.

Once I got the proxy rig done I asked my group if they wanted an overview video of the controls or not. Wasn't sure if they'd be naturally familiar or were planning to play around. They said yeah it would be handy, so I did one.

When modelling the character, I first created a new front and side view inspired by the concept art, which I sent to the group to greenlight the modelling process. Given the time that it could take to create a character I want the group to be happy with even the earliest stages of development in the model, this being the reference in this case.


My first attempt at creating the face went horrible, you'll see that in my demonstration video, but the first startings of the face were poly modelled, starting with a small plane and extruding edges around the shapes of my reference.


Once I had the foundation of a face, I then did some draw-overs, both as reference to help me model the character but also to show to the rest of my group the direction I had envisioned with the modelling in its current state. This really helped me visualise oddities in the model that I'd move on to fix.

I put the facial modelling to one side as it was going really slow and it was clear the character could use a new body. With my new knowledge of Zbrush I thought I'd give it a go. I used simple boxes and the multicut tool to get the right shapes to extrude, the topology of this was awful, but I knew I could fix this after sculpting.

The character with some polymodelled hands, sculpted robes and finally somewhat a finished head. I believe at this point the lower back of his head was still a gaping hole ready to a neck.

I moved on to baking the normals for the robes. Found that arnold has different options for interpreting the Normal map needed to tweak these to yield correct results on the lighting.

Incorrect settings made texture seams very noticable.

The head was looking okay, but I was kind of unhappy with how some of the topology looked lumpy. I tried to consider ways to smooth the topology but found that I didn't know of anything in maya.
Knowing the Zbrush handles meshes in a clay-like manner I took the model into Zbrush to smooth, I soon found myself sculpting over the original model and going between Zbrush and Maya cleaning up some of the unnecessary topology. Overall, the head that came out of Zbrush looked much nicer without the lumps but having looked at the previous character head for so long, it felt as though this new head was a different character.

I was skeptical whether the group would like the changes and assumed maybe I'd made a mistake in terms of the art direction. They all loved it. So I rolled with it.

After working with the new changed mesh, and looking back on the old one, its now clear which is the better model.

His floating stubs needed some work, so here's some feet.

UV Unwrapping Face, Hands, Legs. *Robes unwrapped earlier since needed for Normal Map

First pass textures, these were just block colours so that it was clear to whoever would work on the textures where seams are and various elements like eyebrows and the parts of the body and how they associate with the UV's. I created a displacement that I wasn't using for the character in Maya, but found that it worked well as an AO.
(of course, since then I now know how to bake AO as its own map)
(I'm new to this okay)


Received positive feedback for the texture work.. which is oddly satisfying when all you did was quickly scribble down information as future reference.
There's been some changes to the textures since this first pass, but it's generally pretty similar.

This is where I should have added the ears to the character but instead I decided I'm an idiot and I'm going to rig. (How did I even completely overlook ears!?)

Rigging process begins on the face. The overall skinning of the character went fairly smoothly but it was soon found that the long sleeves were a challenge to rig and not entirely worth the effort of trying. so I left them rigid for the time being.

Some minor hiccups that my group encountered, initially very confusing but once I found the problem it was an easy fix.

I then moved to rigging the face, an interesting task that I was expecting to get stuck on more than I did, I'm thankful for the facial rigging tutorial that I followed over the summer break and a lot of the information came back to me fairly naturally. I kept the functionality fairly simple with enough to get a wide range of motion, but the character is not well equipped for lip sync. I'd soon like to develop this character to be capable of lip sync in future.

Now this handsome bald child with no ears is ready for animation..
yeah, there's a few things left I suppose. I wasn't too concerned about the hair as my plan was to rigid skin to the head, the ears I am a little sad about as I wont be able to get nice deformations in the skin but all works well.



Definitely some work needs to be done on my ability to do hair sculpting but overall fairly happy with the results compared to the normals.

To close off the character, here's a little demonstration video, final textures and a Sketchfab view

Out side of the character I also worked on the lighting in the 'Shrine' room. Trying to keep to my original concept art.

A mix between hard and soft lighting so that we can get the important shadow shots with crisp shadow outlines.

Furthermore on the lighting I created the Candle Light in both the scenes using a simple mesh, skinning some joints and applying a mesh light so that it would emit. Making this an infinite loop and keying the exposure to add a slight flicker effect, which can be seen in some shots.

I didn't take much part in the environment as we already had plenty of modelling going on there already.
(Totally not because I am really pedantic about keeping scenes clean and upkeeping naming conventions and the horrible mess that my group call an environment doesn't bug me at all)
In all seriousness, we did have a lot in that department and I trusted them not to break anything. I also wanted to stray away from micromanaging the group. there's definitely a lack of consistency in the way people work but I've been looking past that.

I suppose I also did some animation. so, here's all of that in a nice breakdown reel.
Will note that a lot - if not all of the early proxy model animation is not mine and is only there to serve reference to development and the base animation I had to initially work with. All non-proxy animation is mine however.

This project has done wonders for my skill set, both on a tangible level and that of mindset, I feel my approach to one of these projects will be much stronger next time and my overall ability to get things done in a confident manner has improved. As a group we had little to no major disputes. Whether that is good or bad, really depends. I felt we didn't quite have enough of a disagreement that ideas got changed or pushed forward, there was a lot of silent agreement. On the ideas side. There was much more conversation on how to approach tasks on a technical level but a lot of those disputes ended with one side being a generally better idea than the other.

No comments:

Post a Comment