Ears are evil

Well, trying to model or sculpt ears anyhow. Maybe I'm the only one, but I have always had an incredibly difficult time getting ears to look right... but I guess practice makes perfect right? I've been looking at alot of tutorials and modeling breakdowns to get an idea of how other modelers approach ears. It's not that I don't understand the actual anatomy, I just need to figure out a good way to re-create it in CG.

I found this great reference photo from this awesome forum thread, this modeler has been working on a photorealistic human model for the last year and a half, and watching his progress is pretty amazing

New Approach to Non-Photorealistic Rendering?

My friend just linked me to this really interesting article about a new approach Pixar is working on that will make animated characters look actually hand-painted rather than simply being stylized CG. I'm curious to see where this goes, since one of the debates I'm having with my thesis is how I should make my characters look, if I should mimic Rockwell's style exactly and go for a very painterly look, or bring the characters more into the realm of stylized CG and/or stylized realism.

"Pixar Makes Painterly CG"

Render Breakdown

As promised, here's a more detailed breakdown of my lighting and rendering comparison. It would have been too much of a beast to try to make it work in the blog format, so I compiled everything in a PDF file.

click to view PDF

* If you use Chrome, the file will load in the browser instead of downloading *

Initial Rendering Comparison

So I was finally able to set up a series of render comparisons between Mental Ray (with unified sampling), V-Ray, Arnold, Mantra and Keyshot. This first comparison was a pretty simple set-up, mostly working with figuring out render settings and different light attributes, since each renderer has it's own specific way of setting up lights. Now I get to move on to more material specific comparisons and see how the renderers handle more complicated shaders like glass and subsurface. I will be posting a more in depth write-up about this first comparison shortly.

** to view image at full resolution, right-click and open image in new tab **

Arnold!

So I've been playing with SolidAngle's Arnold renderer and I'm having a blast. It's so simple to use and will get you really nice results pretty quickly. It uses a different method of sampling that lets you keep your renders low quality for quick look development, and then when you're ready to render, you turn up your samples and get a high quality render without worrying about the look changing (ie no need to worry about using low-poly stand-ins or final gather settings, etc). Both Digital Tutors and FxPhd have in depth tutorials on using and optimizing Arnold, but there are some good free tutorials floating around as well. One good one I found was by Sony's Jon Tojek, which manages to cover most of the basics in about 20 minutes. Even though Jon does cover linear workflow in his tutorial, I found another video that is a quick and simple explanation of setting up linear workflow in Arnold, for anyone who doesn't want to scrub through Jon's video to try and find the right section.

Jon Tojek Arnold Tutorial

Linear Workflow In Arnold