[Aardwolf]

I don't get why this exists. I'm not saying it's a bad model, or that it won't be (possibly) better than the original or the one Galemp had, but a model having collision errors doesn't mean you have to completely remake it. It might be as simple an issue as flushing the bsp cache and reconverting it.

Also, (personal preference) I like the more retail-like shape (x-lattice) over the blobby tubes.

[Enioch]

I ran the previous model (by Galemp) through a cleanup program, and it turns out it was a manifold (I found the right word! ) mesh, which means certain messyness in game meshes. No matter how many re-conversions, there would always be a part of the mesh where the collision 'normals' (for lack of a better word) would be inverted, leading to big collision problems.

This model is non-manofold. Any (if any) collision problems can be fixed by re-exporting/ re-converting. And it's got more polies (now at any rate -the modeled tanks knocked the polycount up to 14k tris)

[Snail]

And it's got more polies (now at any rate -the modeled tanks knocked the polycount up to 14k tris)

Not necessarily a good thing given they have similar amounts of detail.

[Enioch]

Not necessarily a good thing given they have similar amounts of detail.

They don't. The aft section is more detailed on this one.

[Galemp]

The main hull, with the pipes and greebles, is much more interesting here. I especially like the texturing job--the red painted striped on the fin are inspired.

I don't know what you're doing with the gas tanks--it seems like a ton of duplicate work and to me they look a little disproportionate--but I wish you the best of luck.

[Hades]

Where are the teeth?!

[Bobboau]

that was sympathy.

looks much nicer, but I really think you should put the 'teeth' in, it was one of the more noticeable details of this ship and it was one of the things I first thought to look for when this model was mentioned, the lengthwise running panels at the front of the ship are very distinctive and play a significant role in the ships character.  and it is only a texture mapping issue, you don't have to actually model them in.
IMHO


you should add some forward facing antenna arrays coming off the lower boom thing.

[Enioch]

@Bobb & especially Hades: Jeez! Relax... I'll get the teeth done, if you insist, but we'll meet halfway: you'll have to settle for a texture job. I'm not going back to the mesh at this point. And I'm not going to change it right now; I'll leave it for later, when the tanks are finished, textured and tested in-game.

@Galemp: Thanks a lot. I did think the red stripe was a nice touch, if I may say so myself. It's nice thatsomebody agrees.

And yes, I figured a quicker (and cleaner  ) way to do the tanks, without any complications; sleeping over a problem does help. I'll scrap the current sub-object and start work on another.

As for antennae in the fin: there are some, but they're thin, untextured and they don't stand out (yet). As soon as I chrome and brighten them up, they'll be more visible.

That's all for now, I think.

P.S. Oh, BTW, Bobboau, there are a few questions I got for you, concerning the animation codes. Can I PM you?

[Bobboau]

unless the admins have done something wierd with your permissions, you should be fully capable of sending me a PM, though if it is a technical question it might be better done in public because you might not be the only person with the question. and it's been a few years seance I worked on the animation code.

[Enioch]

Right.

We seem to have encountered a leeeeetle problem...Or not. See below:

A single tank 'panel' (3 tanks, each next to the other) consists of a minimum of 260 tris. I can't reduce them any more, without the smoothing getting shot to hell and back. Which, after some hurried calculations gives us:

For the sides: 2(both sides)x3x5x260=7800tris
For the bottom: 5x260= 1300tris
For the aft top: 3x260=780tris

Which gives us a grand total of 9880 (say 10000) tris. Just for the tanks. So the whole mesh would be around 2k tris. Is this acceptable?


And @Bobboau: My question is this. Can I have a continuously 'swinging' subsystem? For example, can I have a radar dish rotate 60deg at the beginning of a mission, then stop, then rotate back 120 degs, then stop, then rotate back 120 degs and so on and so forth until the end of the mission and without resorting to any extra scripting, but using only PCS2 and table entries?

[Galemp]

Take a closer look at the Zef I posted. Put it in FRED and fly around it if you have to. I only modeled the outside surfaces of the tanks, and cleverly filled the inside with a single trapezoidal prism.

[Bobboau]

I would say shoot for 1k, but 2k will work ok.

as for the animation you wanted, I did at one point have it possible to insert a small script into the table for animation purposes, but I don't think it can be done without any sexp or scripting.

[Enioch]

Ooops...typo. Not 2k. 20k. 

And I realize how you did it, galemp. That's how I'm doing it as well. Did you really think I would model the whole tanks? I'm only doing one half of them. And if they were cylinders, they'd be 8-sided, but now they're half-cylinders, so they're 4-sided . Still, the smoothing looks acceptable...

And, again, sleeping over a problem helped. I think I can knock off about 3k tris, just by using another configuration for the tanks. I did me some hurried calculations in my noggin', and I think that I can reduce the total tris for the tanks to about 7k.

So the whole of detail-0 (without any turreets ) would be 17k tris.

It's a lot.

But tell me what you think anyway...

[Raven2001]

Can be 17k for Lod0... the tanks could be detail boxed

[Bobboau]

ok, those tanks should only be like 2200 polies.

im figuring 21 polies per tank 3 per panel, 7 panels times 5 rows

[Enioch]

Actually, they're less.

The new configuration works like a charm! The port and starboard tank 'packs' are each 1.4 k tris, which brings the total polycount to ~13000tris. Still need to do the bottom and aft tanks (again, separate subobjects) but they wont bring the polycount over 14k tris, which I think is quite reasonable for a ship of that size and a model of such detail. 

Still have to texture them, though. I'll probably use another, small (512^2 or 256^2) texture for them. Galemp, if it's OK with you, I'm thinking of 'stealing' yours... 

Renders up shortly


[attachment deleted by admin]

[Galemp]

Still have to texture them, though. I'll probably use another, small (512^2 or 256^2) texture for them. Galemp, if it's OK with you, I'm thinking of 'stealing' yours... 

Well DUH. That's why I put the texture in there.

In BMP form.

With a template for a normal map.

[Black Wolf]

Looks good to me - 20k polies is high, but detail boxing will be your friend there. The textures are brilliant 9but please, don't put any teeth on it, that'd be ridiculous).

[Enioch]

Who sez anything about 20k? Did I say anything about 20k? I don't think so...

I think I'll stick with this model version (see att.)

The front tanks (port, starboard and bottom) are 3.5k tris and the rear (changed, improved and retextured!) rear is ~700tris.

Which gives us a grand total of 13.3k tris (for the whole model). I think it's fine. Don't you?

About the rear: I considered the fact that the docking port (possibly for pumping out the gas) is there. So, I deviated a bit from the canon model and textures. I put no tanks there. Instead, I did an internal submodel with a pipe matrix. There are two 'big' pipes (which are a continuation of the two other big pipes at the rear) running down the entire length of the glowy part and twisting up to end under the first 'square panel' where the dockpoint will be. And then there are a few smaller pipes, barely visible, snaking around.

All in all, I like it. 

@BlackWolf: You know, you're the first (with the exception of VA, who liked the textured greebles) who likes the textures as they are.  But they'll change quite a bit before this tub is released...

[attachment deleted by admin]

[blowfish]

Where is the docking point going to be?