Basic questions about modelling for games and especially source engine.

Ok I never modeled for a game bevor so I need some basic info D:

1. How many polys does a average player view weapon model have?
2. How many polys does the third person view model have?
3. Does the source engine accept 3 sided or multi sided polygons or only 4 sided?
4. What do I have to consider to make the animation process as easy as possible?
5. Whats the max or average texture size of the first and third player view model?
6. Is their stuff the source engines hate like specific ways to model asf?

 

Wait, are you going to become a modeller now?

 

I will model some BE guns in my semester vacation for fun and maybee we get something productive out of it

 

1. How many polys does a average player view weapon model have?

I can't give you a straight answer to this as it depends on to which crowd you're catering, but going up to 3k - 5k should be perfectly fine nowadays

2. How many polys does the third person view model have?

Again no straight answer, but trying to keep it below 1k is a generally a good idea.

3. Does the source engine accept 3 sided or multi sided polygons or only 4 sided?

The real question is whether you would need to tri strip your models, and the answer to that is no.
The source engine only accepts tris, but the compiler automaticly triangulates your models, however, it will also automaticly do the smooth shading for tris that do not have it defined. If you want proper smooth shading, it's best to triangulate your model yourself and smooth shade it in your desired 3d program, or at the least go down to quads and smooth shade yourself, or else you'll end up with messy shading, especially if some of your polies have steep angles that would split up the shading.

4. What do I have to consider to make the animation process as easy as possible?

A rig, learn everything you can about forward and inverse kinematics, blend shapes, deformers and proper workflow on skeletons and keyframe animation.

5. Whats the max or average texture size of the first and third player view model?

Most of the time you'll see one texture for the body and one for the face, for good detail it's best to go with 1024 or more for both textures.


6. Is their stuff the source engines hate like specific ways to model asf?

Tons of stuff, you'll need to first figure out how well your 3d package will export for source, for instance maya has very little support, the exporter hasn't been updated in the past years and theres a lot of work arounds required to get things to work. Even if you're going for XSI there will be a lot of work arounds you'll need to be familiar with before you get things to work.

 

Thanks

 

About texture size, for the view model, fuck it and go with 2048, because seriously, we don't live in the P4 age, we live in the QUAD AGE BABY

 

I HAVE A PENTIUM 4 BITCH

:p

 

Not everyone can afford to upgrade their computer every 4 years, especially in this economy.

 

great. I've barely gotten the hang of making the bloody model. Forgot all that export and dump into Source part.

.....

Also, I stink at texturing. hardcore.

 

Too bad for them.

Make it look as good as possible, looks first, polycount second, although there is also such a thing as efficient modelling, so make it as low as you can while retaining the shape you want, but don't worry about adding more in to make it look better if you can.

 

You can always down-res something. It's much harder to up the poly/pixel count.

 

You can use high-res texture and MIP mapping will take care of "low detail" version of it.

 

No it is just as easy.

Also this, for you people with a desire for modeling:
http://boards.polycount.net/showthread.php?t=63496

 

It is?

I find it much easier to select and edge loop or ring and hit chamfer than I do to manually weld all the points back together again after I've done it.

Texturing is easier to downscale than vice versa, but not modelling.

 

I agree. I'm not all that experienced, but if I use the reduce polycount function I usually end up having polys that look like shit and I'll need forever to make it look good again.

 

Make the downscale from Highpoly to Lowpoly manualy. Tada no problems.

 

Except for the part where that takes forever.

So as I said, it's easier to make the low poly first than then do the high poly.

 

Yeah it takes forever but the end result is better (if you know what you are doing).
Imo to many people think that their must be a auto button for everything in
every "creative" programm from photoshop to 3DsMax.
If you know what you are doing the manual way produces often the better result.

 

So, manually easier and better to go hi -> low.

Auto button, does better job low -> high?

 

Backwards.

Harder to go low to high.

Easier to go high to low.