Archived

This topic is now archived and is closed to further replies.

modeling technique

This topic is 4960 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

When I model I usually just create basic shapes (spheres, boxes) and move the verticies and extrude bits until it looks right. This usually works well for me. Or you could use another model as a template.

Nukes rule 24 7

Share this post


Link to post
Share on other sites
Polygonal versus -- What? CSG? Raytracing?

You''ll find that the industry pretty much entirely uses polygons, or in particular trigons. Why? Because they are very fast to render; Creating a scanline rendering engine for a triangle is amazingly simple and very very fast when done in hardware. Higher-order primitives can be down-converted into triangles without too much trouble before being sent to the graphics pipeline. (And this can be done transparently by your video card''s driver.)

Having said that, however, I tend to use mostly quads in my modeling. To me, they''re a bit more natural and easy to use instead of dealing with the raw triangles. (And as long as I am careful to keep them co-planar, they convert to trigons with no muss.)

---
- 2D/Pixel Artist - 3D Artist - Game Programmer - Ulfr Fenris
[[ Gaping Wolf Software ]] [[ GameGenesis Forums ]]

Share this post


Link to post
Share on other sites
quote:
Polygonal versus -- What? CSG? Raytracing?

He is not talking about rendering methods, but modelling.

There are sveral ways to model, most of which you will find in the most popular 3D graphics applications (Softimage, Maya, 3DSMax, ...) which include direct poly modelling, nurbs, subdivision, etc... all of which provide advantages and disadvantages.

quote:
which 3d modeling technique most game developers use?
is it polygonal modeling? why?


I doubt it would be polygonal modelling at this point. Look at most upcoming games, which claim 10000+ poly models. Modellers would be hard pressed to tweak thousands of vertices in any reasonable amount of time. They will more likely use resolution independent modelling techniques (such as nurbs and subdivision).

Note though, that the game will most likely internally use polys since that is what a graphics card can process.

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
my method

1. Get reference materials (pictures/sketches of what your modelling),and import them into your 3d software. Usually done by applying them to a box as a texture/map

2.Begin modelling.. use box modelling (extrude bits from a box) for characters and primitive modelling (boxes/cylinders etc) for objects.. Its just easier to texture characters using box modelling , generally you add a meshsmooth after texturing to smooth the model...you can texture after the meshsmooth for greater detailed textures if you like...

thats for ingame.. cgi is whole different story
nurbs is over complex and hi-poly, plus box/polygonal modelling is the accepted way to model ingame...

Like an artist i try to learn many techniques as each suits different problems better.. i actually use splines to make buildings... in my 3d class polygonal basically was the first method and splines the 2nd method...

Share this post


Link to post
Share on other sites