More Animal Robot Concepts

Here are my three. I tried making a biped horse to see if I could get around the concerns about animating quadrupeds. I think it looks pretty goofy.

Concept - Animal Robots

Here are three sets of concept drawings of various animals found in Europe, with the exception of that raptor. Each set contains the actual animal study and an overlay of mechanical design. The design is quite rough. The purpose of these drawings is to get a sense of creature's silhouette. Once we have our creature, I can then move on to create the actual functional mechanical body. I was hoping Europe has more scary animals that you know will mess you up if you mange to meet one in the forest. very disappointed.... I guess we will probably end up with a mishmash of many animals.

First Set [Greyhound, Goat, Lynx, Bear, Fox]

Second Set [Wolverine, Raptor, Spanish Fighting Bull, Deer, Wolf]

Third Set [ Badger, Mink, Beaver]

Be sure to let me know if there is a particular animal you are interested in seeing

Cat in Question

So at out last meeting, we came to the conclusion the cougar is out of place. The setting of our film leans heavily towards a European style village, and as such, the cat will not work. We also decided having a fully mechanical creature would better serve the purpose of our story.

But we are still tossing around the idea of having the mechanical creation we come up with be based loosely on an animal. The animal would have to native to Europe, something the creator (mechanic) would take inspiration from.

Now I have come to accept that the cat will probably go, but I was still curious about what the feline population is like in Europe. After all, it never hurts to look into possible ways of salvaging old work, so long trying to save it does not skew your perspective. The most important factor is how well it serves the story after all.

I found one interesting possibility. A lynx. Here are a couple different breads and where they originate from.

Iberian Lynx




I really like this guy. He has that really menacing look to him that I feel gives him a lot of character. This breed is on its way to extinction and lives in Spain.

Carpathian Lynx

There home is in the Carpathian Mountains of Romania.

Eurasian Lyxn

pretty close relative to the Carpathian Lynx from what I can tell.

Now of course that doesn't take out the difficulty of animating a quadruped, but in my opinion, a quadruped would be a whole lot easier than something that say moves like a spider. Of course 2 legs would be the easiest to animate, but from a design standpoint it is the most difficult to create something unique.

We'll talk more, Saturday at ten. Everyone try to bring in 3 different designs.

Quadruped Rig Test

Here is a Quadruped Rig I made and tested out today. Still have a lot to work on though. Tired to put him in some poses that I thought he would be in during the short.

Walking on a roof scoping on the solider on the ground

On the prowl.....

RUN!!!!!!!

Stalking...

My New Mel Script Creation

After many sleepless nights and missed meals this week, I am finished programing the holy grail of time-saving mel scripts. It takes every selected Material and creates Matte Passes for them, organized in a way that is extremely easy for the compositor.

In short, we now have control over every object in post production.

More Details Here

So what that leaves on the compositing front is the ability to create custom passes with Deex's system as well as gamma correction for linear workflow (complicated subject, I'll talk about it later). In good time.

The New Materials We Will be Using!

As I mentioned a while back Maya has some big problems with Render Passes. Namely a 5X Render time increase bug (which I've notified an Autdesk rep of) and lack of EXR image file compression. I have been taking a very active role as of late in the Maya Rendering community at CGtalk to help solve these problems. After many months, we finally have a solution courtesy of Deex!

The Standard Shader

The Skin Shader

(+Many More)

The Render Outputs

Unfortunately there is only a Maya 2009 x64 PC compile at the moment. We'll need to find someone who can compile it for mac os.

So what does that leave? Well, we still need an automated Matte Pass creation solution. I'm trying to write a script for that here that will do it for us, but I'm no programmer so it may be a while before it's ready. Hoping someone else will come along and help.

Either way, this is a major breakthrough for us today. I'm hoping once semester starts we'll be able to start building assets with these shaders right away.

Cougar WIP

Here is the cougar/cat I have been working on. As you can see, I still have a long way to go in terms of detail, so before I go any further with this model. I want to see your thought on what I have so far. I also have some questions for you guys that will help me improve the design. I have a particular hard time modeling the back end, since there are still many unknowns. Your feedback will greatly help my process. : )

Do you see the cat and his mechanical parts exist in the same world setting as the girl & soldier? If not, what makes you feel that way?

Will we be having close-up shot of the cat? This will help me figure out the level of details I will need to achieve.

Are there things you want me to simplify/fix for the rigging process?

Are we expecting the cat to be fully rigged when the semester starts?

Another Shot at the Girl

So I really need to start modeling the girl's clothing soon if we expect to make the deadline of having the fully rigged and ready character before semester starts. Wanted to draw one last proposal before our meeting tomorrow. This is a Photoshop paint over of the 3D model, so everything here is proportionately accurate to the model I did months ago. I did it this way so I could get an idea of how old she looks, and get a closer preview to a final result. I also deliberately made her hair shorter and took out a long dress because it can cause big headaches trying to fix dynamic simulations.

To me she looks in her early to mid teen years, which I am fine with. May not be quite as young as what we may have been imagining in terms of story, but let me explain why this age is ok with me.

To start, it is incredibly difficult and time consuming for me to make age corrections. Sure, it's an easy fix drawn out, make the eyes, forehead, and cheeks bigger, and scrunch the mouth and nose closer together, but with 3D comes another complication. I have consider what I like to call the porciline doll effect:

An example that comes to mind is a Christmas Carol. The children look well, like porcelain dolls. To me it is a bit out of place in contrast to the other characters, but it's done for a reason. It's not just Zemeckis' movies either. Take a look at this new trailer for Toy Story 3. Notice Andy, as a teenager, and the children look like they could literally be from different universes.

There is a trend with 3D films of greatly stylizing children characters for a reason, more realistic children look downright awkward and creepy. Watch Polar Express if you are not following. So for us, if we choose to go even younger I would have find the happy medium between a less awkward stylized look and a look that fits with the other characters and world. Quite frankly I'm not so sure it exists.

In conclusion I would much rather have a slightly older girl that fits in, than a younger child that either looks awkward or feels out place in the universe. That is just my opinion however, perhaps you feel differently.

Ugh, I Know, More Lighting

For the past few days I went back to testing out even more lighting possibilities, seeing how it's one of the bigger bottlenecks impeding production. If you haven't been able to tell yet, this stuff can get pretty complicated.

After reading through a ton of Siggraph Papers about the spherical harmonics solution and learning there is no automated implementation for Maya/Mental Ray we can use, I happened to stumble upon this:

Smart IBL

This quite literally seems like a gift from the heavens. It works with Final Gathering, the most commonly used method to calculate indirect light. The problem with the Final Gather approach is normaly you get results like these:


YUCK!!

The ways to avoid this flickering is either to increase samples to an insane amount, which still can cause problems and increases render time, or significantly decrease the samples and blurring them together which causes a loss of quality.

But thanks to Smart IBL, they have an automated solution that gives awesome results in less render time. Here is my quick test:


YAY!!

No noticeable flicker! There is a little variation on the walls, but that is from the light bouncing off the cubes which should be there.

Machines & Some Quick Sketches

I did these few quick sketches during some brainstorming sessions to find ways to incorporate mechanical parts into our existing feline model. I also throw in some character studies I did.

The first option I came up with is to replace the back portion of the feline with exposing mechanical parts. I want the front to retain the organic feel of the feline, so people will still be able to recognize it as a cat, unless that is not what we want.

The second option is to simply have inner mechanisms exposed on parts of feline's body to suggest the cat is mechanical.

I do not have a sketch for the third and last option, which is an exoskeleton that will be put over cat's body. I do not like this idea because it sort of take away cat's nimbleness and make it into a clunky awkward killing machine.

Let me know what you think. : )