[Eric Lagman]

Had a quick question for anyone that may know. I use surface planes as emitters in solidoworks, but when they get exported out and I open them in studio there are tons of triangles on them. I thought many triangles on emitters=slower render. If so is there a way to avoid this. I need the mesh settings to be high enough so that the other parts of the scene are not faceted, but not the planes. I plan to try a test between an emitter from sw vs a single plane emitter from Cinema 4d and see if there is a speed hit.

[JDHill]

Yes, there will be a speed penalty. Unfortunately, SolidWorks lacks any kind of per-object mesh control.

[JDHill]

Hold on that - I found one way: you need to have your emitter be its own part file. In the image quality settings in .sldasm, there is a checkbox that decides whether or not the settings from referenced parts will be overridden - 'Apply to all referenced part documents' - make sure it's un-checked. I'm not sure if un-checking this modifies the actual referenced documents permanently or not. Also, if you make sure your emitter plane is square, you can get it down to two triangles - if it's rectangular, it will have more than two.

[Eric Lagman]

You da man. Setting the mesh image quality on a per part basis then uncheck the apply to all referenced part documents under the top level assembly mesh setting gets the emitter polygons down to 2 with square surface. The high mesh settings for the objects I want more detailed remain in tact. Wow no telling how much time I was loosing because of high polygons on my emitters. Hope it wasnt too significant. I will run a small test to see the benchmark difference.

[Eric Lagman]

It makes a huge difference. I did a 2min test with the exact same scene. 1 with polygons cranked to max on the whole scene including emitters and one with the global mesh settings off and high where in needed and low on the emitters where I did not.

High polygon scene after 2min Benchmark of 52.02 SL 1.55
Low polygon on emitters only after 2 min Benchmark of 86 SL of 2.55

I wonder how long I have been shooting myself in the foot on this.

[JDHill]

I'd guess that to be somewhere around three years or so >>>

-----------------------
Eric Lagman



Joined: 21 Mar 2005
-----------------------

Ouch! Sorry to say, emitter poly-count is the first thing in the optimization FAQ.

[Eric Lagman]

I checked the studio template I have been using most of the time and the emitters are only two polygons. So only the recent project I have done had this. So its only been a week. Phew. Funny thing is I replaced the emitters with simplier ones that are in the same spot and size just with different mesh settings and I am not getting a faster benchmark. Maybe the other one showed such a difference because it was a simple scene with just a floor and some lights. Whereas the scene I just updated and expected a higher score is more complex geometry and has displacement going on.

[Mihai]

Hey Eric,

It's not always good to go simply by BM. In the more complex test you may not see much difference in BM alone since it's complex already but judge instead by the noise levels after a certain time. With the more complex emitters you should see a noiser render.