Faces versus Nodes

If any of you have been in my classes, read any of my books or seen my videos, or followed any of my ramblings in magazines and on the forums you'll know that I tend to rant and rave about modeling efficiency.

One of the prime factors slowing productivity in most of the offices I visit while training is just too many unnecessary faces in the models.  That's not to say that everything has to be as low polygon as game modeling for example, but there's just no reason for having lots of faces that don't serve a specific purpose.  Optimize, optimize, optimize!

However, there is a related topic in the total number of objects in your scene.  As far as I know each object in 3ds Max or VIZ is stored in memory as a node and there is minimal, if any, difference in the node space of a complex 3D object or a simple 2D line.  Therefore, the more objects you have in the scene, more memory needed to store them.  It follows then, that if you can combine multiple objects into a single object you'll be more efficient.

For example, a scene with 1000 objects and one million faces is much less efficient than a scene with 100 objects and one million faces and your render times can be noticeably longer, especially if you're working with limited RAM resources. 

The fix is easy

You can combine multiple objects into a single object with the Attach options in the Edit modifiers or in the editable state or you can use the Collapse utility.  The attached objects become Elements at sub object level and you can always select them and Detach them into single objects again with unique names.  You can also apply materials and modifiers, and (in Max) animate the Elements at sub object level, so you aren't limiting your ability to work with the objects.

There's no good reason not to work efficiently and if you build your models procedurally you can adjust the density of your mesh objects if you find that you need more detail for whatever reason.