Having a lot of difficulty :/

[Downsider]

Recently made a very nice pistol model in 3ds Max 9.

Can't get it to MDL. Can anybody help? I mean, I've tried literally everything. Tried 3ds Max -> SMD -> Milkshape -> MD3 -> Quark -> MDL but when I tried importing from Milkshape, certain vertices got unrigged off the model.

Jesus Christ it's impossible. HELP.

[MauveBib]

There's an md2 plugin for max called QTip, but it only allows exports of models with tiny skins that you'll have to resize afterwards in qme or something.

md2 > mdl is trivial, save in qme, load and resave in quark.

[Downsider]

Unable to place all textures in specified skin size
NOTE: This error is often caused by a texture being tiled excessively in a model resulting in necessity for a huge image.
Go to http://www.qtipplugin.com/faq.htm for a complete explanation of this error.

FAQ doesn't tell me anything I wasn't already aware of and have tried.

oO Wtf mate, wtf.

EDIT: 64x64 works, eh, but I need 256x256. I'll figure something out. Major thanks.

[MauveBib]

Yep, shrink the skin down to export, then resize it up again in qme.

[Downsider]

Well, this doesn't exactly work. When I export from QME 3.1, and chuck it in FitzQuake, it says the model number is <insert gigantic number here> and it should be 29.

What.

[frag.machine]

29 ? You mean something like "BSP version is <number>,not 29" ?

Well, you should never get such message with alias (.mdl) models, since default .mdl version is 6.

[Downsider]

29 ? You mean something like "BSP version is <number>,not 29" ?

Exactly my thoughts. It says LoadModBrush as well. However, it's not the brush model; it's quite obviously the alias model. In Darkplaces, it doesn't give the error, but it does replace the model with the diamond-shaped placeholder.

[MauveBib]

For some reason qme's md2>mdl convertion is a bit bust.

To fix it you have to export from qme as mdl, then load it in quark, then save as mdl, then load it back in qme to import the skin, then finally save as mdl again.

Silly, but it works.

[Downsider]

For some reason qme's md2>mdl convertion is a bit bust.

To fix it you have to export from qme as mdl, then load it in quark, then save as mdl, then load it back in qme to import the skin, then finally save as mdl again.

Silly, but it works.

Now QME is giving me shit about a floating point overflow.

This is real fun and all, guys, but, you know, it's bullshit.

EDIT: Not to say I don't appreciate the help, I do, seriously. I understand that it's a whole project to get this working and many people have their own little methods. Just venting here, more or less.

[MauveBib]

Try resizing the model without changing the view angle? Too large models are all I can think of that cause floating point errors in qme

[Downsider]

Try resizing the model without changing the view angle? Too large models are all I can think of that cause floating point errors in qme

It's definitely not too large.

[Dr. Shadowborg]

For a possible alternative try the blender model exporter mentioned in this thread here:

http://forums.inside3d.com/viewtopic.php?t=1210

[Downsider]

Find me a way to go Max to Blender without losing anything (UVMaps, animations, rigging, bones) and you got a deal.

[chris4268]

Maybe you should try importing to Blender with different extensions, Mr.Breadsticks.

[Downsider]

Well 90% of that post makes very little sense to me but thanks anyway, mate.

EDIT: Oh, do you mean extensions as far as different model formats? Yeah, I know that much, but very few preserve all aspects of the model including bones, uv, etc. Collada seems to be winning so far as far as supported features go, though.