I work in Lightwave 3D on a dual-processor 2.5 GHz Power Mac G5 that Jim Cameron gave to me for the purpose of building my CG models. Even that computer, though, has difficulty dealing with those high polygon counts. When I tried an animated fly-thru of Titanic's boiler rooms, I could render only 10 frames at a time, else the program would crash.
I am aware that my models are too polygon-intensive for animation, but I use them for other purposes. In my Marconi Room model, each wireless component can be disassembled like the real thing. In my Turkish Bath model, each of the screws of the coathooks can be removed from their countersunk holes. In the boiler rooms, each rivet in the boiler casings is individually modelled. I need this level of detail because I use the CG models as forensic tools, to brief divers on what they should find and catalogue what they have found. In this way, for instance, I was able to tell by comparison that the stoking indicators in Britannic are of a different style than the ones used in Titanic. I also can learn how an apparatus works by building it from scratch in a virtual model. I learned about the wireless apparatus by building each component, so much so that I was able to fill in gaps left unintentionally in the period documentation.
You will never see my CG models used in an animation (although I did compose a brief animation that re-created John Chatterton's dive route in Britannic's No. 6 Boiler Room for the upcoming History Channel show on Britannic), but I can render single-frame images that can be used for illustration in books or on TV. Some of my Marconi Silent Room single-frame images were used in the un-aired Cameron edit of the Discovery Channel's "Last Mysteries of the Titanic," and I will use them as illustration for the book that I intend to someday write about the Marconi apparatus.
I taught myself how to use Lightwave and the only way that I know how to build an object is to create every detail. I am not expert in using image, transparent and bump maps to give the illusion of detail. I cannot build the entire ship -- as you have done -- because I would create too much detail in the model. For my sinking simulation, I borrowed someone else's model. I wish that I could build the entire ship, as you have done.
I also wish that I worked in Studio 3D Max. When I worked on "Ghosts of the Abyss," the animators used Lightwave and it was from them that I learned the art of computer modelling. Nowadays, most animators working in film use Studio 3D Max. That means that for the fly-thru animation of the Turkish Bath Cooling Room in LMoT, the animators at Earthship Productions had to build their own version of my CG model, using orthographic views that I rendered for them. If I had Studio 3D Max, they would not have had to copy my effort...I could have shared my models with them directly. But, I work primarily on a Mac and Studio 3D Max does not run on a Mac.
Thanks for the information. I'm amazed that you can do it even in a program like Lightwave, on such a computer. The computer is definitely a powerful one, but in the PC class still.
> Jim Cameron
Wow. If I ask nicely will he give me one too?
I've seen quite a lot of your CG work, including the Marconi room & other stuff in GotA. I was thankful for it, as it really brought to life that part of the ship. Superb book and DVD, btw.
I wish I could model in high detail like you're doing, but as you pointed out, a whole ship would just be too heavy like that. Can only be done with the computing power of a company like Pixar or ILM. How you model (whether high or low polygon) is determined by the purpose of the model which is of course decided at the outset.
As I said, our interior the meshes will be about 10 times lighter than yours, but even that would to too massive for a whole ship, so it won't all be render at once. Groups of rooms will be rendered when you're in them, then turned off when your in another part of the ship.
We can only look forward to the day when supercomputers are as commonplace as the PC is today. I think mulitcore processors are probably the way forward, but instead of dual, 32 or 64 core etc.
Yes, Cameron gave me the computer and I gave him all of my CGI. He got the better part of the deal, since I spent two years researching and building the million-polygon models. No one, not even Cameron, could afford that effort if I charged the industry rate for the time I spent. Even so, the deal worked out well...I got a computer that I couldn't otherwise afford and he got CGI that he couldn't otherwise afford.
The CGI in GotA was done by Lightship and Creative Logik Universe animators. I learned how to work Lightwave while advising those animators. For the Marconi Room model in the GotA companion book, I built some basic 3D objects in Solidworks (my first 3D program, before I bought Lightwave) and Ken Marschall added detail in Photoshop.
For LMoT, we tried to export my Lightwave objects in .3ds format so that the Earthship animators could use my objects in their Studio 3D Max program. However, there were conversion issues that the animators could not solve, so we had to give up that idea. I have tried importing .3ds objects into Lightwave, but they don't import well, even with the proper plug-ins.
We should have this conversation on CGTalk...probably boring the Titanic people.
P.S. I just looked at the Marconi Room model in the GotA book. That version is very much out of date. I continued my research after GotA and the 2005 dives gave us new information. My latest version of the Marconi Room is significantly different from the GotA version. It's also much more detailed, since I learned more about 3D modelling while I worked on the Marconi Room model in Lightwave.
this model is absolutely amazing! you all do such a great job with 3d modelling, i only wish mine could compare. i took a 3d animation course this semester and learned how to use maya, but i have yet to be able to properly tackle anything on the titanic.