Why do most modellers display their models as egdemapped quads instead of triangles? Since this is game development, triangles are the building blocks.
When viewing quads, there are two ways the triangles can be aranged, and in low poly modeling that can have a big difference.
Just wondering why..
Major Clod: Interesting that you say alot of modellers use spline or nurbs to model, I've only come across one, everyone else has been editable mesh/poly modellers.
Just curious how many people use these ways of modelling? I'm a max/maya poly modeller predominently even when its higher poly models.
Redwyre: Its just when artist pimp work to their peers its more about the 'look' than the efficiency, we of course find efficiency important, its still a major part of a modellers role but doesnt make others go, oooh aaaahh. lol
Hmm... I guess maybe I have that impression because I learnt a lot of my skills from tutorials and people that used spline type techniques. Although I originally started to model with editable mesh tools, I've found myself using spline based tools more often to create my initial mesh... especially if it has a lot of curved surfaces. I don't use editable poly that much at all.
I do all my stuff with edit poly in max5.1. Can't beat it. I find with splines, you can't control the density of the mesh in areas where lots of parts meet like the fingers, groin and armpits.
As for why people keep things in quads. It might have something to do with the way 3d engines light the models. If your trianges arn't arranged in a roughly quad like fashion, then it can create lighting artefacts. This is mostly apparent on large smooth areas like the chest and thighs.
You can test it by animating a light to spin around your model in the viewport and watching how the light is picked up by the polys.
:)
Building in spline cages? Nurbs? Similar high poly tools?
no no no we're still living in low-poly worlds with low-poly characters...to reduce a polygonised model from nurbs et al is futile, and twice as much work.
Simple primitives are first made, then extruded, bent, warped, vertices welded etc to conjoin primitives. Guaranteed low polys, plus simplicity and low mistake / corruption factor. btw The default settings for Max and Maya are quads with hidden interior edges.
This is also beneficial to game artists because the maps they create in Photoshop are usually square, or rectangular, and seeing the applied map and extent of UV's on a quad is clearest with a hidden dividing edge.
Simply put, the inside edges are mostly of no use.
A lot of modellers create their meshes using spline cages, nurbs or other similar higher poly tools, because it is easier to create smooth, flowing surfaces than with working from an editable mesh from the start. Of course the result is transferred into a normal mesh at the end, but the edge inside the quads is left hidden. I find when working on a model of around 3000-4000 tris, its less messy if you leave these edges hidden. When you display your wireframe mesh, leaving these edges hidden just looks better :) I can assure you that most modellers will check that all of these edges are oriented well though.
Also, leaving these edges hidden makes it easier to use Meshsmooth later on, as it relies on quads to smooth the model properly.