Possible r_speeds issue with build 19

[FourthX]

On one of my current projects I tested the map in 3 different configurations, build 19, build 14 and old quake2 dp.

Quake2 and build 14 show about 750 in one spot while build 19 shows over 1100 in the same spot. the whole map overall gets 2-300 more tris in build 19 compared to build 14 and old quake2.

Can someone else try this out and see if they get similar results? Check your map in build 19 and any other version you have handy and compare r_speeds.

FtX

[y00tz]

Tip: a good way to compare is to record a demo and type/say your r_speeds for specific still-shot views, so that when you  try it in another build, you can watch the demo and you'll have something to compare.

[DrRickDaglessMD]

Is reflective water turned on the build 19 version? Possibly the engine is drawing some water, possibly out of shot. With reflective water it would make it render the scene twice, possibly explaining the hike in w_polys.

Dagless

[FourthX]

Nah reflective water was turned off along with caustics as soon as I 1st saw the inconsistency, the issue remains.

I'm going to go with what q2 and build 14 are giving and assume I have a glitch in b19 or my config is wrong maybe.

FtX

[y00tz]

<dumbQuestion> Is your FOV the same in both builds? </dumbQuestion>

[FourthX]

Yes, in all 3 I set the fov to 120.

FtX

[jitspoe]

Old builds don't count the water polygons.  The quake2 style water is broken up into a bunch of polygons in order to create the wavy texture effect.  These can cause significant slowdown, yet weren't counted at all, so I added them to the count.  I also added the polygons rendered for reflections in the water.  Build 19 will give you more accurate r_speeds.

[FourthX]

Bummer :/, I liked seeing 950 better than 1100. In no case is the map laggy or slow so I guess it doesn't really matter.

FtX