Monday, May 9, 2011

3D rendering 101 - or - how to lose your mind in 10 minutes



We've been getting a lot of comments about our 3D renderings of HORNET lately, with many people asking who did those for us. When we set out to produce those graphics back in January, we knew we would need professional help (as if we didn't already). So we took the old CAD drawings I used to analyze the ship waaaaay back in 2009 and passed them on to a really (and I mean this) first rate rendering firm. Just over two months and many, many emails later, we were at a total impass. In their defense, a sailing ship is no easy thing of which to understand the multitude of complexities - even enough to draw one effectively. Sails were floating in mid-air, and the rig was a complex tangle of virtual lines. Masts raked insensibly in every direction. We were getting nowhere fast.

And so in typical SWO/NHS fashion, and with three weeks remaining until showtime, I set to work figuring out what I would need to beef up my old CAD program to create something in the way of what I was looking for. In searching 3d rendering programs, the cheapest start out at around $900 - far more than our budget allowed. We had emergency reserve funds, but I didn't want to tap those unless absolutely necessary, and I was SURE it was what we needed. In the end, I found a program called Blender. This powerful software package is what NASA and other agencies use, and comes with an unbeatable pricetag - FREE. So what's the catch? Well, it has arguably the worst user interface ever devised. The learning curve was steep, but I had to give it a shot.

And so, after a whole week beating my head against a wall, I began to make headway. I actually started with the sails and rigging first, knowing they would pose the biggest obstacle as I had absolutely no idea where to start. I then began work on the hull, deck structures, guns and other details. I didn't make everything, mind you - I cut a good many corners by purchasing ready-made items such as the 3d carronade, 3d ship's wheel (which I edited to match Hornet's) and flags (many of which still aren't done). And night after night, until all hours, I was glued to my computer screen working node by node, edge by edge, face by face until I had produced a presentable 3D model.

Bear in mind, all the while we have four other teams working on sub-projects that I have a hand in - legal working on the project charter, setting up the corporate sub-accounts, filing paperwork with every government agency under the sun, and building Mini-HORNET. I was losing a lot of sleep.

In the end, what did we get? The rendering is halfway decent but not great - there are many points that still elude me about the actual rendering process and setting up 3d environments. But what we did get is a few volunteers who stepped up and began learning the program. Now, I'm looking forward to handing off the project to some more tech-savvy volunteers who will work on producing better renderings and even some animations! But there is a certain feeling of satisfaction - not only in knowing that I met the deadline when everyone said I was crazy (again, I agreed from the get-go), but in that others are picking up the ball and running with it! Charge onward, guys!

NNNN

No comments: