Bean Caught is not officially completed, but we did screen at RIT this past Tuesday and received great feedback and I believe people actually enjoyed the film. This was especially nice considering that for the prior 7 days leading up to screenings, I only had 19 hours of sleep, meaning I was sitting in a chair animation, fixing, shading, rigging, rendering, compositing, organizing, etc for at least 140 hours straight. The leftover 9 hours went to eating and showering. :o)
Sometimes I cannot believe how much work can be pulled off in 20 weeks at RIT, but it’s a lot. The film my awesome partner and film co-creator, Meghdad Asadi Lari, and I produced was full of dedication: from the day we conceived the idea to the very last minute of submission. Even a trip to the hospital that Meghdad unexpectedly took didn’t stop him from essentially arriving back and hopping back onto the computer – though I told him he’s crazy not to rest and get healthy again.
We also had a couple other MAJOR developments that were unexpected. The last week of production, before we had to complete our film for a strict screening deadline, we discovered after shading our referenced character rigs that when re-imported back into the shot files, nothing worked… None of the animation worked! Vertices were everywhere, shaders were applied to some objects while not to others, etc. It – was – a – night – mare. I almost cried. However, after convincing my buddy that I could create a script to fix it all, and that I wasn’t going to waste the rest of our time to do so, in less than 2 days I came up with 2,000+ lines of code to save our project. The MEL script that was developed allowed me to select the rigs in trouble and automate a process that would apply the shaders within the shot files instead of in the reference files. And to be smart about it, everything about the script and even the shaders were modular – meaning that if updates needed to be done, they could be implemented easily and would allow for quick additions to code without sorting through meaningless, uncommented, spaghetti code. I also created a shader file, comprising of all our procedural textures for our characters, to allow for easy updates and linking… because I didn’t want to shoot myself in the foot if we needed to adjust something for all 30 shot files.
The other slight crisis occured two days prior to screening deadlines. Our clean-up animator sent us back files where the references where removed, but the geometry and rigs were baked right into the file. Render settings were all messed up, shaders were not working or were simple checkerboards, etc. So, with a little patience and not too much cursing, I was able to modify my 2000+ lines of MEL, and with 55 more lines of code, I was able to create a work around that allowed us to still use the animation and simply re-reference the environment scene files. Needless to say, we got back on track though it delayed us enough to question if we’d ever make the deadline.
So yes, luck was definitely on our side, but it was also a great experience for me. I’ve had to code on a whim to help or save projects in my previous job, but never have I plugged away for 2 day to save something at this magnitude! I’m just glad I was able to pick up MEL within a very short period of time create several powerful tools that saved us lot of frustration and time. On that same not, just in the past several days, I also created another MEL Script to handle shot re-renders as we’d like to put out a production quality pass on our film before releasing to the public.
I’m anxious to put out our piece. It’s our best work yet and it seemed to get a lot of good responses from a variety of people that night. I just hope things get easier and better from here… though a lot of great lessons were learned! :o)