I don't know where I read that, but it amused me and stuck in my head.
Hi! I haven't been updating the vfx blog because I've been crazy busy. Work has me deep in MEL and Python code most of the day, writing software to manage animation data, shuttle it this way and that and process it in interesting ways. There's also the secret project that everyone isn't talking about but is quietly and excitedly waiting for (assuming they have any idea what it is). We're on super-special-secret-lockdown at work, I'm guessing for the rest of the month. I can scarcely go to the bathroom or grab a soda without passing through at least two security doors now.
Ok, I admit it, we've been contracted to create the lost tapes of the lunar landing. Seems NASA has gotten enough flack about the original landing with people claiming they can prove it wasn't real that they've decided to create more proof to put the concerns to rest.
Or at least that's my amusing cover story until I'm allowed to talk about the other thing.
Between all the other little tasks, I'm in the process of working up a new tool for moving complex animation data around the pipeline. The idea is to be able to read and write an XML file that contains as much data as anyone would care to put onto a model. The model can be updated, have new elements added to or removed from it, all sorts of parts of that model can be animated at a whim by the artist and passed along to subsequent versions of the model, low res or high res versions, keeping all of the data intact, allowing for scaling of the animation data if it's transferred to a different model entirely, even passing along worldspace location and orientation of parts of the model (and maybe even eventually the full animation) to other platforms like Houdini and Realflow.
It's a little important for what we're doing on The Curious Case of Benjamin Button, but the existing tools I developed are working ok there (even if they're still a little limited and occasionally awkward - they're getting better on a weekly basis). There's already a small test of what I'm doing in a broader sense: we have a particular element with a moderately complex rig on it that I'm publishing and subscribing the animation data on.
The Big Goal for this tool will be on T4 which we should be working on soon. That'll be a huge show and we've got some really epic-scale work on it: most likely we'll be needing the more sophisticated animation tools for managing our data throughout the pipeline - so that's what I constantly keep in the back of my mind as I script the current version of the tools. They need to serve the current project in the most complete way possible, and they need to be able to expand to do more than I'd currently dream of asking them to do.
It's a lesson that's come up recently because of a little tool I wrote that I've mentioned here before "CurveBaby" - CurveBaby allows me to take a scene with an animated camera and one or more animated objects, preserve the camera's relationship to one of those objects but change that object's animation - or, conversely, reanimate the camera and have the animation of the object automatically adjust to still look the same from the camera. It's a powerful tool, and one that I can barely imagine how I'd get through half a day without using it at least once.
Big tools make big projects go a lot more smoothly.
