Adding Cross models and images to encyclopedia

[Vehek]

I don't know which enemies are specific to disk 2, but disk 2 battle models don't seem to be in the pack. Is it hard or annoying to dump them?

[FaustWolf]

Here's a dumper MDenham made followed by some .dlls that may have to be included for it to work:

http://www.sendspace.com/file/ytxsan
http://www.sendspace.com/file/qpknfq

I'd feel comfortable with calling MDenham into the discussion if we can't figure this out (I seem to recall we had trouble getting the Terminus Traductions tools to work properly with Disc 2, but maybe I'm wrong) but I'll leave that decision up to ZeaLitY. Now that we know how to work with the file pointer table at the beginning of each disc though, I bet we'd be able to find the missing Disc 2 enemy models without too much hassle. Especially if the dummy files on Disc 1 correspond to actual files on Disc 2.

Other than getting the raw data, we'll have to go through the process of getting the textures in TileMolester since we don't have a conversion tool for that variation on the .TIM file structure.

[ZeaLitY]

Go ahead and let him in.

[FaustWolf]

Done. Not sure when he'll pop in next.

Sorry for the holdup on the field models, I got carried away with Chronotorious. I've still got about 40 folders to go through and then arranging them neatly. I'll upload an archive as soon as I've finished.

[MDenham]

Done. Not sure when he'll pop in next.

Was in yesterday before work (stupid 3:30pm-midnight shift) but didn't really have anything to say at the time.

Mostly just going to hang around until assistance on my part is actually needed, rather than clogging up this thread with unnecessary commentary.

[FaustWolf]

Ooh! Ooh! I'm not sure what the status is of the file dumper I believe you made like a year ago, M. I think there was one you made that successfully dumped everything from both Disc 1 and Disc 2; I could have sworn we were having problems with that using the original Terminus Traductions tools. We might need that capability again to get at some Disc 2 models we're no doubt missing.

[Vehek]

Transparency doesn't seem to be working for me. Is it working when you try?

[MDenham]

Ooh! Ooh! I'm not sure what the status is of the file dumper I believe you made like a year ago, M. I think there was one you made that successfully dumped everything from both Disc 1 and Disc 2; I could have sworn we were having problems with that using the original Terminus Traductions tools. We might need that capability again to get at some Disc 2 models we're no doubt missing.

I'll have to rewrite it as I lost the source to it (it wasn't that hard to make the change to the existing Disc 1 dumper, though).

On that note, a Subversion server would be nice so that I can readily keep the source available for other people to look at (for obvious reasons, preferably not directly associated with here).

[FaustWolf]

Hold on M, is the dumper linked in the following post yours, and does it contain all necessary components?
http://www.chronocompendium.com/Forums/index.php/topic,7902.msg174783.html#msg174783

[MDenham]

Hold on M, is the dumper linked in the following post yours, and does it contain all necessary components?
http://www.chronocompendium.com/Forums/index.php/topic,7902.msg174783.html#msg174783

That should be it, yes.

(I'd still like to rewrite it so that rather than just the generic progress-bar type of result it currently gives for Disc 2, it shows what's what on Disc 2 instead.)

[utunnels]

Yeah; it's been an issue with importing since the beginning. It's been too long for me to remember if some lines like that are endemic to the characters in-game, though...

LOL

Code: [Select]

uv[0] = uv[0] / float(textureWidth - 1)
uv[1] = uv[1] / float(textureHeight - 1)
It seems changing line 770 and line 771 in the python script to lines below will fix the problem.

Code: [Select]

uv[0] = uv[0] / float(textureWidth )
uv[1] = uv[1] / float(textureHeight )

[FaustWolf]

Has anyone tried recompiling with utunnels' alteration? I'm not sure if Luminaire's popped in here yet.

[Luminaire85]

I tried it briefly.

The advantage of utunnels's version is that your texture polygons always have UV coordinates that are exactly between pixels. The disadvantage is that you are not able to use the last row and column of a 256 x 256 texture, even though there is data there. Not sure which is better, although I'm tending to agree with the utunnels version.


I've summarized where I recall things being in the two attached images. They depict the same exact Serge model viewed in the normal Blender editing window and with the render previewer. As you can see, neither image is perfect.

The normal view shows a thin line down Serge's back and some unexpected orange splotching on the backs of the shoulders. This seems to be due to some blurring of adjacent pixels, and ensuring that the UV coordinates fall exactly between pixels does not help here. This goes away in the render preview. (Although I've not shown it here, the render preview also allows for proper handling of transparent portions of models, e.g. Glenn's bangs.)

However, the render preview shows a problem at the bottom front of Serge's shirt. This is due to the existence of two faces occupying nearly the same space. One of these faces points toward the viewer, while the other points in towards Serge. Only one of these faces should be viewable at a time (based on the location of the observer), but for some reason the renderer is trying to draw them both.

It would be better for ZeaLitY if we could fix the render view, as this would allow for much larger images to be taken. This means fixing the double faces issue. Ideas anyone?

[utunnels]

I'm not familiar with how uv works.

I tried to export the bulb model into 3ds format (using the original script).
The left most vertex have a u value of 1.0000, but it seems that is beyond 100% of the texture width(the snapshot below).

So in those 3d renderer, are u/v values start position of the pixel, but in CC they stand for the pixel itself?

Edit*
Forgot the picture.

Edit*
It seems in CC, uv ranges from -1 to maxsize-1?

[FaustWolf]

utunnels has reported that the Hydra textures and Black Dragon textures were accidently saved as 24bpp PNGs as opposed to 8bpp PNGs. If that's preventing importation into Blender, make a list of any more you find and I'll reup the model pack.