I don't know where to start.

When I'm writing these blogs I usually go through the images and renders I've saved over the week that I want to share and base the writing around them, I didn't used to do that, but they've become so long and sprawling and infrequent that its the only way I can remember.

So if you'll please turn to figure 1.0

Here we have an image of Moss Scott, the mascot for Kernel during studio time. He partly exists due to the number of times I asked Ryan to cover everything in moss, partly as a nest for the rivet we found in the studio (I think) that matches Leonard's rivet perfectly, and partly to keep spirits up with his jokes.

Figure 2.0
This is the look of the final knowledge cloud and its thanks to Kai's fluid fanciness that it looks this detailed, I borrowed a combination of his and the script's lights to light it like this. He's also placed a particle system in the center which orbits the central light like a nucleus. This was the first time I'd properly used lights with a negative intensity before. Its great because you can actually subtract light from the surroundings and its what gives the cloud that deep red underside.

Figure 3.0
This is the shot that the mask was made for, to properly reveal Leonard, but that strap has been a complete bastard to simulate. Not only did it take forever to fix as it intersected with his head and face over and over again, when it came to render time Maya just ignored the cache and re-simulated it, managing to get the strap right through his nose and mouth for the last 50 or so frames. Why?
I love the red and blue colour palette of this shot's lighting. What was useful about this shot was it forced me to make some final tweaks to the fur and eye shaders so they held up to close scrutiny a little better.
Figure 3.1
Not what he signed up for.

Figure 4.0
I've taken to lighting Leonard alone at first in some scenes now as they are so heavy with geometry and dynamics that it takes half the render time just 'translating the frame' which essentially is it copying everything to the RAM I think, so it can render. This way it copies relatively little and I can start seeing what changes I need to make without have to wait 5 minutes.

Figure 4.1
I got carried away and starting compositing this one. Delicious. You can just see the beginning of Liam's lights and the animated shader I worked on with him. It should look spectacular in motion.

Figure 5.0
As Len's darkest moment where he realises his oxygen is broken, I figured it was time to take that darkness quite literally. And also time to break out the lens flare, but not too much, keep it classy.
Whats weird is this was rendered with final gather, which seems to have covered up the sub surface effect I was getting through his ear. I believe this is because the diffuse light is stronger than the scattered light, but I don't understand why there are such dark bits in his ear. I may have to fake the final gather to get that back, I'd turned it right down anyway so its only go to speed up render time.

We have around 5 weeks left. Its going to be interesting. I am still optimisitic.

Pixel Propaganda

I'm proud to say I understand 90% of what they're talking about in the new release of mental ray's update notes.

An article I'd been meaning to read for weeks turned out to be pretty interesting. Its on one of Pixar's cinematographers. She talks about how her painting allowed her make some observations that helped improve the atmospherics lighting in one of their films.

A good interview with the father of the Metal Gear games. he talks about why and how he innovates and some of his feelings about his own creative successes and failures.