So here’s a quick recap on our 48 Hour Film Project experience:
Yup yup yup. Long story short: we completed 4 minutes of 3D animation in Blender in 48 hours. And it came right down the the wire, too. We turned in our film, rendered and ready to rock with a mere 10 minutes remaining. And here’s the really remarkable thing: we started off this year with even less than we did last year. This year, all we had when the project started were two partially complete character models (they were missing their mouthbags, otherwise known as mouth interiors). These models weren’t rigged or textured or anything. Furthermore, we didn’t have a set or any props modeled, either, and a lot of our crew wasn’t available for the full 48 hours. Here’s the result, with a couple bits of polish, rendered in HD (and below that is some “making of” information):
It was due to the huge bucket of talent and dedication in our team that this was able to be finished. We have some absolute machines in this crew. It’s incredible how much work was done in such a short period of time. Everyone pitched in; often doing more than a single person’s fair share of work. The interesting thing is that animation didn’t really start until late, late Saturday night. An incredible number of frames was cranked out in that short period of time. I can’t say enough how honored I am to have worked with all of the people on this team. Not only that, but I’m still surprised that they actually agreed to participate in such a ridiculous idea.
To get an idea of just how hectic the weekend was, have a look at these timelapse videos that Hal Dowdy took while we were working. They’re each about 8 hours worth of footage crammed down into one minute of time. It’s pretty amazing.
There are a couple highlights that I’d really like to focus on here, though. One of them is Bassam’s new Rigamarule rig-retargeting script suite. It took Bassam roughly 5 hours to rig our Round character. However, with Rigamarule, the same rig was retargeted to Square in about 20 minutes, most of which was dedicated to cleaning up vertex weights. Without Rigamarule, we would’ve started animating much later and probably would not have turned the short in on time. Bassam just released Rigamarule 0.2 on the blog for his Tube project. If you’re a Blender rigger, I highly recommend that you check it out.
Besides our talented crew, the other element that made this production possible were our renderfarms. In particular, I’d like to mention the Animux renderfarm that Mark Puttnum runs as well as the very cool people at BlenderFabrik. Without these two farms and the small team of “farm hands” to wrangle them, there’s no way that we would’ve gotten this thing done on time. The best we’d be able to pull of would be OpenGL previews… and who wants to watch that?
A somewhat surprisingly helpful piece of technology for this project was TinyChat. Since most of our real-time production discussions happened via IRC, it wasn’t a huge production tool. However, it was great for morale. Since most of our team was located elsewhere throughout the world, it was an incredible help to be able to see everyone else working (and freaking out as the deadline loomed closer). It was also a great way for the outside world to watch our production as it happened. We actually recorded most of the video from the TinyChat session, but I haven’t really gone through all of it. Multiple cameras at over 48 hours of footage it going to take some time to look through. I promise to eventually post something, though.
All in all, I had a great time. We finished. It got screened with the other 48 Hour films, and we’ll find out the results of the judging this weekend. I’m not expecting too much. We’re at a bit of a disadvantage by doing animation for a challenge that was designed for live action filmmakers, but I feel that this was an incredible accomplishment. I can’t wait to do it again.