Timeline for Why do game engines convert models to triangles instead of using quads?
Current License: CC BY-SA 3.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 6, 2020 at 9:06 | comment | added | Simon F | "but a triangle cannot be split into anything else than triangles." is actually incorrect. A triangle can be converted into 3 quads. See computergraphics.stackexchange.com/a/9480/209 | |
| Feb 1, 2013 at 13:06 | history | edited | Nate | CC BY-SA 3.0 |
deleted 1 characters in body
|
| Jan 29, 2013 at 22:53 | comment | added | Hatoru Hansou | +1 Because is an answer I can safely use as reference when somebody ask me the same thing, to my bookmarks. Only say that, I always have thought that the reason triangles are the small possible primitive is because with the imprecision of float point arithmetic, tris are the only safe polygon you can guarantee to be planar in all cases, with quads you cannot guarantee they will be planar all the time. Modeling software probably show objects as quads or n-gons as convenience to the modeler but apply transforms/render dividing polygons as two or more triangles. | |
| Dec 20, 2011 at 23:39 | comment | added | Suboptimus | In OpenGL ES, as part of the effort to simplify the API, quads and polygons aren't even supported. So a practical reason to use triangles is because you don't have any other choice. But the reason a streamlined API like ES avoids other primitive types is due to reasons outlined in this and other answers. | |
| Mar 28, 2011 at 16:39 | vote | accept | Grant | ||
| Mar 9, 2011 at 1:25 | comment | added | Grant | Awesome! That makes a lot of sense when you think about it. Thanks for the answer! | |
| Mar 9, 2011 at 1:24 | history | answered | Nate | CC BY-SA 2.5 |