Preserve quad meshes?

Started by archz2, January 16, 2022, 12:34:45 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

archz2

Everytime I import a model with quad faces, keyshot splits it up into triangles which increases the face count tremendously. When dealing with high polygon models, this causes a major issue as my Keyshot hits a max limit of rendering polygons in GPU mode and I have to resort to CPU mode instead.
Sharing some screenshots of a sphere model that I imported in Keyshot with different quad face count.

Here's a sphere with a count of 400 faces


which become 760 in Keyshot.



A sphere with 1909 faces


that become 3818 faces in Keyshot!








What can I do to preserve the quad faces as they are?

TGS808

Nothing. KS converts anything that is imported into triangles. Unfortunately, there's no work around you can do to change that. It seems Luxion would have to change the method in which KS treats geometry when it's imported.

archz2

Thanks a lot for replying. I should definitely start learning another render engine now. I was in a dilemma earlier but now your reply has made me more confident to take the leap. It's really difficult when a geometry with original poly count of 5-6 million polygon becomes 12-14 million polygons and I have to resort to CPU rendering.

Re-meshing then zooming through the geometry to check for quality losses consumes lots of precious time leading to loss in productivity.

Morgan


if it is interesting have a look here, I have the same problem with this solution
https://forum.keyshot.com/index.php?topic=28318.0

Eugen Fetsch

Every render engine will convert polygons to tris. A quad is always made by two triangles and it's just the matter of visualization in the viewport. Even the software you're using makes tris, but shows and counts just quads instead. So what you in KS is the number of tris (always the double amount of the quads). 6 million quads become 12 million tris, 1909 quads become 3818 tris. The "panel node" doesn't show this statistic.

mattjgerard

Read this from a forum about the Saturn system- "There are plenty of reasons to prefer triangles. They always exist in a plane, whereas four vertices can have a third dimension. A triangle always subdivides to another triangle, and a new polygon can always be created by adding just one more vertex to two others, whereas a quad can be subdivide into a quad and a triangle, while you have to add two vertices in the same plane to create a quad. Every other polygon can be expressed as triangles, but quads can't express all polygons - how many quads do you need to express a triangle or heptagon? To create triangles, you have to overlap two quad vertices which is making it a triangle! Triangles are easier to work with in the real-time engine, more mathematically 'atomic', and focusing performance on pushing triangles means accelerating all geometry to the same degree."

Basically it sounds like triangles are the smallest plane object that a surface can be broken into, and rendering in quads doesn't solve anything, it actually introduces a possible third dimension which is a hinderance in calculations, mapping and coordinate systems used deep in the file systems. There isn't anything you can't do with triangles, but there are things you can't do with quads, so rather than having 2 systems of surfaces, they abandoned quads to keep things simple. In the beginning it was all about poly budgets and GPU horsepower, and that never really went away. Back in the day, people in charge wouldn't blame the hardware for the inability to render something, it was the artists fault for not being efficient with their scene/object/character. GPU's natively render in tri's and changing render engines won't solve your problem. You need to look at the tesselation of your models, the efficiency of your polygon budget, get more GPU's, or employ one of the many rendering services. All those are much more viable and much less painful ways of solving your issues than trying to learn a new render engine. This is from me having worked with a half dozen in the past, every time i had to / wanted to switch from ProRender/Octane/MentalRay/etc was a HUGE lift. Nothing is the same between them all.

archz2

#6
Thanks a lot for your post Matt. Makes a lot of sense!

Quote from: mattjgerard on January 18, 2022, 06:47:00 AM
You need to look at the tesselation of your models, the efficiency of your polygon budget, get more GPU's, or employ one of the many rendering services. All those are much more viable and much less painful ways of solving your issues than trying to learn a new render engine. This is from me having worked with a half dozen in the past, every time i had to / wanted to switch from ProRender/Octane/MentalRay/etc was a HUGE lift. Nothing is the same between them all.

Currently I'm building my portfolio of rendering, therefore it won't be wise to use cloud rendering services. I'm planning to learn unreal engine 5 as it converts high poly scenes into nanite mesh which makes scenes more manageable when GPU comes closer to running out of memory. I hope things get better.  I want to buy a new GPU but they're being sold at overpriced rates in my location.