Rendition card

[sharangad]

It's coming along.

https://youtu.be/OAvIETekkTs

Thanks to your script I can parse most 3DO files, but some are giving me grief. The 3DO conversion is causing a lot of problems as is the MIP to MI4 conversion.

[sharangad]

Looking a lot better. Some texture coordinates are still a bit off:

https://youtu.be/V02hEvj--vg

[EDIT]
All the issues with the 3DOs was because of a few bugs in the code.

I seem to be having trouble detecting when to scale.

[/EDIT]

[jxeboy]

That looks like its coming along very nicely! Thanks for all the effort you are putting into this!

[sharangad]

Thanks, I've got almost all the polygon's texturing correctly except for a few.

Can't post a video because my isp is throttling me for some reason, having trouble accessing email as well.

[sharangad]

@checkpoint10, how are the textures applied? Some 3DOs don't have any mips, but need to be textured.

[EDIT] My problem is that some 3dos are being modified when they shouldn't be. [/EDIT]

[checkpoint10]

Can you give an example of the problem you are seeing?

3DOs could have a mix of polys that are either textured or untextured. Flavor 1 is an untextured poly and Flavor 2 is textured poly that needs to point to vertices that have the additional <u,v> coordinates for the mip.

[sharangad]

I think I need to handle flavour 4, which selects mips.

How long does a mip selection last? Until the next one or exiting nesting?

I'm dealing with the mips by going through parent 3dos and then moving on to children. Some meshes only have a flavour 4 and there're other polygons at the same nesting level. I guess when exiting an object the mip selection needs to be dropped.

[checkpoint10]

Pavel and Samsepi might correct me on this, but I believe 99% of the time when we are creating objects from scratch, we code the following such that each MATERIAL (flavor 4) is pointing to one specific POLY:

pointer1: MATERIAL MIP="name", GROUP=X, pointer2;
pointer2: POLY [T] <color> {point0001, point0002, point0003, point0004};
point0001: [< x, y, z>, T=<u, v>];

However, I have noticed that some default 3DOs like the car models may be using MATERIAL to point to a BSP function (Flavors 7, 8, 9 or 10) which in turn may point to LISTs or POLYs or other BSP functions but ultimately it will lead to a POLY [T] or POLY. In those cases, I am guessing that all children POLY [T] are using the MIP that is referenced by the parent MATERIAL.

I still think, if a MATERIAL is leading to a plain POLY (as opposed to POLY [T]) it should draw an untextured poly. One situation that might be problematic is if the MATERIAL leads to a POLY [T] but the POLY [T] is referring to vertices that don't have the <u,v>. But that should not be happening unless the 3DO was getting modified from the default coding.

[sharangad]

is it possible that a nested 3DO could be used with a scaled texture and an unscaled one, using a flavour 15 node?

[EDIT] I think your explanation on the material node helps. That seems to be where I've messed up.

I' have difficulty parsing material nodes. They seem to have no polygons but are one of many embedded nodes. So does this mean the next nested/embedded polyT is affected?

[checkpoint10]

Flavor 15 is the definition of the nested 3DO, i.e. where one 3DO points to another 3DO. Usually trackside objects or wheels on a car model:
pointer: DYNAMIC X, Y, Z, Z-axis, Y-axis, X-axis, 1, EXTERN "3do file";

Which file are you seeing a MATERIAL node that isn't directly going to a POLY [T]? Can you share the .3d text file if you got it converted using my converter?