[mtripoli]

I've been doing a little testing regarding the resolution of an object.

I may be going on a false assumption that I need clarification on.

In Lightwave 3D, each triangle has it's own normal - the more triangles you have, the more normals you have, the longer the render time.

In SW, it appears that "faces" - not triangles, have a normal associated with them (I base this on a simple box - "Scene Data" tells me there are 12 triangles - only 6 normals). If this is true, then the number of triangles should not effect render time as it does in 3D apps like Lightwave?

Juan?

Mike Tripoli

[John Layne]

This doesn't answer your question directly, but it's definitely worth a read.
http://www.maxwellrender.com/forum/viewtopic.php?t=7962

[mtripoli]

I saw this from Victor...

SWX - M~R handles things differenyly than Lightwave - M~R... trying to get a read on how things work...

You there Juan...?

[juan]

You there Juan...?

It's me! Sorry for the delay, too busy these days...
The answer to this question is more complex than it seems. First let's talk about the beta version, the one you all have. Of course the number of triangles in a scene ( specially in emitters ) is important for the render times and memory used, but there are other factors even more important : Whether the scene has a lot of hidden places where the light takes a long time to be calculated, whether this is an interior and the light enters throught little windows with shutters, the presence of dielectrics, the number of emitters... there are a lot of essential aspects for render times, even more than the complexity of the geometry. With the new inminent version things are a bit different. Maxwell does not care too much about the complexity of the scene. In fact, the more complexity, the more power of Maxwell is revealed

And about your question of triangulation, it depends a lot of each plattform. In the beggining Maxwell gets from SW three normals and three vertex per triangle. But then it is done a post process of this data to get just the useful info. For example in a cube you just need 8 vertex and 6 normals. The other 6 are ignored because in each face you have two triangles with the normal pointing in the same direction, thise redundant info is ignored. But in a smooth sphere you need three normals per triangle (one per vertex) to get a smooth object, if you get just one normal per triangle perpendicular to it , you will get a undesirable facetted sphere.

Btw, I use my own version of the plugin showing in the SW viewport the triangulation that is sent to maxwell so I can debug and find issues, see the pic below. You can see how the bodies are tessellated and you can adjust the triangulation quality as you said in other post. I guess it could be useful for you, I can add this feature to the public plugin if you want.

Cheers,

Juan

[JesperW]

That looks like an "easter egg" functionality to me. Meaning its a nice to have but it propably should not be exposed in the GUI. (If you want to keep the "non-technical" "simple setup" image of Maxwell, that is )
/j

[mtripoli]

YES, please add this functionality! I looked yesterday (in vain) to find a plugin to do exactly this. I spent an hour or so just trying to coorelate how an STL file might match the OpenGL triangulation.

I did find a program; OKYZ Raider3D that captures OpenGL, but it didn't work with SW2006 and I'm not sure that it actually showed the triangles...

Not to be pushy but how soon could we see this added? Accounting for what JesperW said, maybe it is something that could be toggled on and off? I liken this to "show/hide" control codes in MS Word; I have them on all the time so I can see the formatting of my document; others don't like to see them at all...

Also, does it show just the triangualtion, or does it show the normals that M~R is using as well?

In all respects, thank you for your detailed explanation; I now understand a bit more about how things are handled between SW and M~R.

Mike Tripoli