Final Post of the Term

Gol and Dav from Simon Littlejohn on Vimeo.

It's been quite the ride these past ten weeks. It's had its ups and downs, but through it all I've learned an incredible amount, and gained new skills in every aspect of 3D art and design - be it animation, rigging, lighting, compositing, texturing, or rendering. It was a slow start getting into the swing of animation, but as the project progressed I believe I became both better and more efficient in the process.

Above is an example of how I had to animate characters interacting with each other.
I had to take into account their individual momentum and balance. Hard stuff.

The week prior was spent getting my facial rigs worked out for both characters, put the finishing touches on the skeletons, skin weights, and attributes of the characters, and set the scene up for animation. I also began animating, and had reached around 25 seconds in. Since this past week I have completed the animating process, hitting approximately two minutes of animation that I'm very happy with.

This shot shows how the graph editor is involved in the animating process.
Using it helps smooth out otherwise jerky movement and better mimic natural motion.
I used it a great deal in the animating process.

One of my main goals of this project was to create a character based animation without the use of voice acting. To achieve unique and entertaining characters I had to use only facial expression and body language to convey personality. As I've stated multiple times in this blog, that was where my concern lied, rather than look development and environmental creation. What you see in the animation is rather sparse, but I don't believe it to be distracting. It is enough to serve as a setting for the story.

I spent a lot of time and research into the skin of the big guy Gol. I took a 3 layer
shader and jacked it up to roughly six layers. That way I would get different colors depending
on the strength and angle of the light.

One thing that I thought would really help give personality to the characters was how they walked. Each has a very distinct way in which they carry themselves, and have different walks depending on the current mood they're feeling. I'd say that was the most intensive part of animating, but it was all pretty intensive. It's a toss-up.

This is just a quick shot of the hypergraph, where all the textures in the animation
are referenced. You can easily tell which texture is which and what it is applied to.

A concern that had been raised by my peers is that rigging and animating two characters in one animation is extremely time consuming, and finishing on time would be rather difficult. While it has been very difficult indeed, and taken countless hours getting the characters even remotely ready for animation, I'm very happy that I chose to do what I did. One thing that I never want to do is shoot too low with my ambitions. I'd rather fail at trying to achieve something beyond my reach than succeed at doing something easily within my grasp. However, that's not to say I think I failed. On the contrary, I'm extremely happy with what I've accomplished.

This is another example of how I had to deal with both characters interacting with eachother.
As Dav pushes off from Gol, Gol is forced down and back, causing an opposite reaction. It's some
crazy Newtonian Physics. Woooooo~!

I could have stuck with one character, and come up with a sufficient story that would have required far less time. However, I wanted to push myself and do something that I was legitimately excited about doing. Because of this I've learned more than I had expected in all aspects of 3D, as well as time management in general.

There were a lot of things that went wrong during the creation of the animation. Rigging and skinning the characters took an insane amount of time, and still have problems that I need to work around. Arms have strange stretches and seams, and at one point Dav's head would deform with the movement of his knee. Weird. The ideation and creation of the two character models was also a huge time consumer, as they needed to be fully articulate and deform in the ways I needed them to. There were definitely times where I just couldn't get them to move how I wanted them to without them going insane. There are also a few framing issues, and funky camera angles that I need to tighten up, such as when Gol kicks the rock. Having white fog in the background will help it drastically however, since right now its dark grey on dark grey (with the rock).

The one part that potentially matters most is the story. I tried to come up with something very early in the term so that I could story board as soon as possible. After brain storming I spoke with the class, as well as respected authorities in digital media. Though I was given a lot of great ideas on how to tighten the concept, the heart of the story remained generally unchanged since around week 3. I applied a few tweaks that I believe improve the story. What I am personally impressed with is how close my final animation resembles the animatic that I had spent so much time on. I'm glad I put the time in that did, as it helped me timing a lot.

This shot points out one of the many pitfalls of animating in Sub-D. Each mesh in a movie
starts out blocky, but when using subdivisions, the program averages the distance between
Edges, and smooths everything out. However, it's very easy to overlook a scene that hasn't
had Sub D applied.

The most difficulty I had was when it came time to work on the facial rig. If I had decided to rig both faces, I believe I'd still be doing that to this day, and never would have even gotten to animating. I have a few posts about my troubles in previous posts, but they're pretty thick with semi-technical jargon. In a nut-shell, each different face pose would compete with the rest of the body in how it moved. The more influence a face had, the less influence the skeleton had. Long story short, I had to re-create my blend shapes multiple times. I'm not saying I'm a pro at it now, but I am saying that I dominate it completely now in every aspect of the word.

After this term, I intend to continue with the animation, and make the environment more detailed, and add fog. That was another difficulty I ran into. I intended to create fog in post, using a "Z depth" pass. To try and explain that for the non-3D-savvy, Z depth is added information on an image or "frame" that has been rendered. Each point in the picture has a number assigned to it that is larger if it is further away from where the camera took the picture. However, all my images lost that data for a reason unbeknownst to me.

This is where you set up your render settings. if you look above, there is the "Depth Channel".
You can turn Z Depth on, however, it doesn't actually give you any Z depth information in your image.
Why is that? The world may never know. But I hate it.

Once I figure that out, fog is going in, as well as depth of field, which can also be done similarly using Z depth. Another thing I intend to do is add more detailed textures to the characters. By design they are cartoonish, and won't need a very high level of detail, but I'll add a bit here and there. I also intend to smooth out a few animations that I would like to spend more time on.

As far as animating goes, I found it very fun and educational to animate two characters interacting with each other. I had to account for the posture, body position, and center of gravity for each character, and how they would affect the other. The entire process was enjoyable. It was also extremely challenging, and frustrating. However, most things worthwhile are. As far as what I set out to do from the beginning, I believe I achieved my goals. It's not perfect yet, but it's not far from being something I'd like to show the general public, and at this stage it wasn't meant to be. I'll keep at it over the break, and get it to how I envisioned it would look if I had a full semester rather than ten weeks. Overall, I'm very happy with the outcome and am grateful for the opportunity to spend as much time as I could on it.

Facial Rigs and Animation

This past week was spent re-rigging the face of Gol, and messing around with Dav's facial rig. I had a lot of issues with my blend shapes, which are the individual facial expressions that are then applied to the base mesh. Whenever I would manipulate the shapes, the base mesh would move into the original T pose. The facial expression would still be applied, but the skeletal rig would be completely disregarded.

After some experimentation, I created an entirely new facial GUI to control the eyes and mouth of the face. As you can see below, its in the similar shape of Gol, and makes it pretty easy to animate compared to how I had it earlier (sliders).

However, after I created the GUI, I ran into the same problem that I had before. After some trial and error I found out I had several problems. The first was the input order in which the blend shapes are applied. The second was that the base mesh had been moved, and it's transforms had been frozen after being rigged, which for reasons I don't understand, made things get ridiculous. However, I redid the base mesh, transferred the skin weights, and re-keyed all the facial expressions on the GUI, and now it works like a charm.


As you can see above, manipulating components of the face GUI affect the mesh, in meaningful ways. For example I have the sneer mapped to the nose (maybe not the most intuitive thing every created but it makes sense to me).


Above is an image of all the the blend shapes I'm using right now. I might have to go in and add a few more depending on how I map out each shot of the animation. So far they've been able to give me a good range of emotion, and I may even have more than I'll actually need, but it can't hurt.

Here is another example of how the GUI is mapped to the base mesh. I have some squinting, some sadness, some surprise, and some sneer going on. It's a pretty good time.


This is what it looks like when it's all smoothed out, without the GUI. He'll never actually give this facial expression, but I thought it would be fun to play around with what emotions I could get out of the blend shapes I have, and see if there are emotions I can't achieve.


The shot above is a bit of the behind the scenes look at what goes into creating the facial rig. To the far left is the GUI, with the Set Driven Keys window open to the right. In that you specify the driver (the GUI) and the driven (the blend shapes), and what drives what. For example the translate Y of the eyebrow will make the blend shape of the eyebrow move up and down. The blend shape slider is on the far right, showing how each of those are being manipulated according to how the facial GUI is being manipulated.

Above is a shot of what it all looks like as I'm animating a scene. Nothing too crazy, but it shows what it's like using the GUI in the scene.


I had to spend a little time messing with the eyelid blend shapes of Dav, and get his rig fixed up a little bit with the weight painting of the tunic.

I created a little GUI for his face, though it's far less extensive since he does't show his face to have emotions (how convenient, isn't it?). I just needed something to easily control his eyelids. It wasn't entirely necessary, but it will speed up the animating process, and save me more time than it took to create the GUI.

This las shot shows what its like messing with the GUI and rig while animating a scene. It's pretty neato. Now its just a matter of finishing the animating. I have half of it done, with a mixture of tough shots and simple shots. It's going to be a busy couple of weeks but all the rigging woes are behind me.

Rigs and Carts

This week has been spent getting the rig in working order for both Gol, Dav, and the cart that Gol pulls. The cart was very simple, as its just a bunch of planks of wood, some wheels, and a pole. Not much goes into that. However, the character rigging is rather difficult. Mostly its the skin weights, but I've got it under control now.


I've also been trying to work out my shots, and get the emotions that Gol will be conveying in the animation. I've been playing with the angles, and seeing which ones look the best.


The image above shows how the rig interacts with the character.


These shots show the test animations I've been working on with the cart and character combined. Right now I'm having difficulty with the foot controls, but once that's worked out it's on to full animation.



This final shot shows a combination of both the cart and character controls.

Gol and the Cart Updates

This week I focused on the big guy Gol - getting his rig finalized and working with facial expressions. One thing I had neglected until recently was the cart he's pulling. I also made sure everything was UV'ed so I can move on to texturing for the next week. Below are some images of the progress I've made. The small character Dav is also rigged, but I can't find where I saved the file. }:(
This is me messing with the blend shape contorls the get some expression from Gol. Next up are facial control curves.


This image shows some of the weight painting for the character. With this model I have separate pieces for the body and the tunic, with the skeleton affecting both simultaneously.

This shows a line-up of the blend shapes I will be working on. Since I don't have a separate head, I had to duplicate the entire body mesh, even though I'm only affecting the face.

This is the rig for Gol. Nothing too crazy. Dav's rig is basically identical, only scaled to a smaller size.



Above are some shots of the cart. All that's left to do with it is texture it. It's movement is done with parent constraints, without the need of any rigging. Hooray.

DVD Menu Images

Big guy Gol's almost complete body Plus More

This week was spent creating the main character Gol (working title). He's a big hulking yet timid fellow who doesn't know his own strength. Below are some in progress shots of the process I took to create him.








I also created the mask for the second main character Dav (working title), which is pictured below. It needs a bit of work, but the basic shape is there, and the topology is ok for the most part. By next class it'll be good to go.


Next up will be taking pieces of Gol and reshaping them to represent the body type of the smaller character Dav. I'll also refine the topology of the mask, and add a shirt to Gol.

DVD Menu Mockup