A Dilemma
After a long series of attempts to try to get around a few problems we’ve experienced, we’ve determined we just can’t get around it anymore. The original models are just about too messed up to fix, in any sort of reasonable timespan. Bear with me.
Here’s an example.
When the Archangel Model is converted from .pak, to .obj, using the *only* converter there is, then from .obj to .3ds, then imported to 3dsmax, it looks like this:
You need one note, for context: The red is *errors* in the geometry. By contrast, this is what a model looks after I remove the egregious errors:
Much less – but the peg was one of the easier ships to remove errors from. The problem continues: Here’s a top-down render of the Pegasus.
Looks ok, right? Look closer. Behind the cockpit, the texture is misaligned. There is a chunk missing out of the cockpit glass. Did you remember that the Peg has *two* sets of scoops? One set is broken, so doesn’t even show up!
This is what it looks like, when you fix the geometry errors. So, basically, to fix any geometry problems, you have to break the texture of the model, then rebuild it.
In the Arch, it’s about as bad as it gets.
I can get it down to about 85 errors. It breaks a whole bunch of the texture, ends up gutting the engine section, and cuts a big slice out of the bottom end. But, i can fix all but 85 errors. It took me about a week to get it to that point. We have WAY too many models to be doing that to. The models are NOT good enough to justify spending that much time on them.
They are ALL like this. Every single one. We have at least 25 fighters slated for release, and to have to fix them ALL that way? It’s crazy. We’re going to have to basically rebuild a few of them.
On top of that, we have capital ships, bases, satellites, gun platforms, etc, etc.
So, I’ll cut to the chase…
We’re not fixing these models. It’s crazy to waste all time on a low-poly circa 1999 model set, with major, major geometry problems. So, we’re going to change our focus.
We’re going to replace them. Build our own. Model them from scratch.
Which, unfortunately, is going to take a LOT longer time. However, the end result will be MUCH more in line with visual effects like this:
I’m sorry we’re going to have to delay – but the end result will be something much, much more satisfying – visually, and technically.
It’ll be worth the wait, we promise.
~ RK
Leave a Reply