Maxwell Render | Official public forum

Maybe this has been mentioned but I think it would be great if you could tell Maxwell Render to apply a mesh, at any specified density, to a displacement object at render time. this way I do not have this high poly meshed model to work with. I imagine it would have be done through the plug-in for each modeling app.

Unless I am missing something, I know this can be controlled with precision but from what I understand the object still has to have a mesh correct? And I am not sure how an mesh at render time would effect uv texture maps.

Aaron

The mesh doesn't need to be high poly to work well. It just needs to have fairly equally sized polygons (no long thin triangles). But then you will need to use higher precision (which is creating a higher poly mesh).

I don't use FormZ, but in Max, I sometimes create a low poly subdivision surface and apply a turbosmooth modifier that only affects the mesh at render time.

I see. I thought I read, not sure if it is true, that having a tight mesh with low precision renders faster then a big mesh with high precision. Do you know if this is true? I have been using tight meshes with lower precision. I just was finding hard to see my model with all the walls of building tightly meshed you know. I will us a larger mesh with higher precision.

Thanks,

Aaron

I tested that theory when displacement was first introduced in Maxwell. Surprisingly, larger polygons and higher precision was faster. Maybe it's worth testing again.

Asmithey wrote:a tight mesh with low precision renders faster then a big mesh with high precision. Do you know if this is true?

Super correct.

tom wrote:

Asmithey wrote:a tight mesh with low precision renders faster then a big mesh with high precision. Do you know if this is true?

Super correct.

I'm still going to test this!

You will discover a huge difference in efficiency, not even close to be comparable.

tom wrote:

Asmithey wrote:a tight mesh with low precision renders faster then a big mesh with high precision. Do you know if this is true?

Super correct.

So we better subdivide a lot the mesh during modelling?
There always some trick to learn...