View Full Version : Question about "efficient" modeling?

01-13-2004, 05:46 AM
I get the idea that lower polys can translate to lower rendering times, but I'm confused by how you know or learn when you have too many polys, or the wrong kind of polys, etc.

Some of the models I see that are put up for critique look efficient to me, but a frequent comment is "too many polys"

Are there any tutorials which explain how to get this sense of how much is enough?


01-13-2004, 06:17 AM
There's no hard and fast rule - it depends entirely on the project.

For real-time and games work you're limited by the target platform and the graphics technology used by the game. For video and film work - well it just depends...

If you're creating an object that will be seen up close it'll need more definition than something way off - if you're creating for D1 then it'll probably need less detail than something for 2k or 4k film work.

It's also a matter of style - some people prefer to keep their models simple and let texture mapping create the illusion of detail - personally, I prefer to model most of the details (I just think it lights better).

In some situations you may even need to create a number of "level of detail" versions of your object so there really is no answer to this question - whatever works for what you want to do - no more and no less.

It's worth taking the time to get to know how the tools and shading of your 3D software work - for instance, when using sub-d's you'd be surprised how little geometry you need to create a seemingly complex object.

Other than that I guess it's just down to experience (and even old pros make mistakes from time to time) ;)