I couldn’t get into the labs today (er, yesterday), so I jumped back into my non-photorealistic rendering tests. I stayed up way too late working on it, so this entry will be short. I promise I’ll have a more in-depth update soon, but for now you just get images. Overall, I’m happy with my progress:
I’m using two different shaders to achieve this result: One for the etching, and one for the color. I broke them up because I felt I had better control when combining them in post. For the second image, I added an ambient occlusion pass to see what it might add.
That’s all. Time for bed!
!!Update!!
I like the non-ambient occlusion image above better, but I’d like your thoughts on both. Also, here’s it in motion:
!!Update 2!!
The earlier examples were a bit rough because, due to version issues, I didn’t have access to a lot of my files. Now that I’m back in the labs, I decided to test the shader on an animation cycle for my crowd sim project. You can see how it holds up below:
And with motion blur:
!!Update 3!!
This is probably the final update for this iteration of the shader. I wanted to see how it would hold up when added to a full environment. Here is the “out of the box” result:
While waiting to defend my thesis, I started a side project! It will be a 5-7 second shot of a crowd of baby bird-egg-dudes running around. The main goal for this project was to revisit Massive and have fun doing so.
The characters and environment have already been created, as well as all necessary animation cycles. I am currently building the brain for my agents, and should be able to simulate the crowd within the next day or two (if all goes well!).
Here’s the most recent WIP of my environment:
Environment WIP
I got a little particle crazy (NUUUUUKE) when my friend suggested that I put a few in the scene, so ignore those. Also, I will be making a few other minor adjustments to the lighting and shading. The middle ground and foreground are pretty bare right now, but my plan is to finalize the set-dressing once I finish the crowd sim that way I’ll know where the characters will be going.
My thesis is in its final stages, so I’m finally going to start updating my blog with CG! First up is a test with non-photorealistic rendering. There is plenty of inspiration out there which is driving my interest, but what really motivated me was the recently released Diablo III cinematic. The begging of it has a really awesome style to it:
So, since the labs are closed for the day, I began experimenting this morning and came up with this:
I’m using the facing ratio of the object to drive a ramp which selects the color. To break it up a bit, I averaged a bump-mapped shader with the facing ratio, and connected that to the ramp. I got the idea for the approach from http://sunnykharbanda.com/blog/2009/10/behind-the-toons-part-2-the-shader/ (thanks!).
I altered the shading network as I saw fit (which I’ll describe once I make more progress), and with some minimal comping, came up with the above result.
It isn’t quite where I want it to be, but it’s a good start. The next step is to adjust the contours to get something similar to brush strokes.
