Does 'poly-modelling' render faster than 'box-modelling w/ boolean/proboolean'

[coolhand78]

just curious to know if using poly-modelling techniques ie. swiftloops and bridges, maintaining quads etc, to create houses renders/behaves faster in 3ds max/VRay than say
creating a box and using boolean/pro-boolean to 'cut-out' windows etc...?

apologies if it's a dumb question...

[Macker]

I much, much prefer the geometry you're left with when you poly-model things - and so does the Vray displacement! It really does make life a lot, lot easier.

[joconnell]

On a flat surface it won't make much difference but if you're got things like overlapping / coplanar faces (two polygons lying on the exact same plane) then it'll cause any raytracer to freak out, that'll e what causes glitches in your renders or slowdowns.

[coolhand78]

it's a 'conversation' i'm having with some at work, i'm arguing for poly-modelling, but not all think it's the best way to go, simply as they dont know how to do it,
so i'm trying to determine what the best points are for arguing my case

[Richard7666]

As has been said, on the V-ray side, displacement will play nicer with proper geometry. On the 3ds Max side, deformations you might wish to apply typically won't work on non-quad geometry.

What exactly are the benefits of *not* modelling in quads? Speed I guess but the loss of flexibility is often not worth it. (admittedly I do often do it, but I also know it's not best practice). I only do it for basic objects such as walls and ceilings if I'm in a hurry.

[Macker]

Show them this https://vimeo.com/80204209 and ask them "what's so hard about that then?"

[Neilg]

Show them this https://vimeo.com/80204209 and ask them "what's so hard about that then?"

skip 19 minutes in first... I wish he'd turn grid snap off too.


I'd probably quit my job and change careers if you took shift drag edges and bridge away from me.

[im.thatoneguy]

The problem with Boolean modeling is that it will often generate long narrow triangles. Long narrow triangles can be kind of a pain for acceleration structures to handle efficiently since they can traverse large sections of a mesh and reduce the ability of the acceleration structure to group them together. It shouldn't be a huge difference but I would expect proper subdivisions to be best accelerated. However, again that should be a pretty negligible difference.

[coolhand78]

Show them this https://vimeo.com/80204209 and ask them "what's so hard about that then?"

it's this video where i pretty much learnt how to do it...