Talk:VIS optimization

Sweet Job on this Tom.--Gear 19:12, 16 Sep 2008 (PDT)

Quake and visleaves

I can't find anything on iD Tech the Quake engine(s) and visleaves. Are they actually Valve's invention? --TomEdwards 18:53, 25 April 2009 (UTC)

I'm not sure. I did run across several detailed docs and tech publications a while back here when I was looking around the QuArK forums, although I haven't looked through them myself. There's also id's Licensing page. If all else fails, you could try emailing someone at Valve or id for specifics. --brandished 19:57, 25 April 2009 (UTC)

Occluders

I've been doing some heavy performance tests in my experiments with occluders, having at least 3-4 visible at a time. During these tests I've noticed next to zero ms cost for calculating occluder culling. So I'm assuming here that occluders by those tests have magically gotten cheaper, but perhaps we never noticed it over the engine moves.

This also coincides with the fact that I've noticed quite a decent amount of occluders in maps from TF2 and L4D2. Obviously considering L4D2 is on an entirely different engine branch though. However for now I think it is safe to say that occluders are fairly cheap, thanks to whatever changes valve made in the system from Orangebox and beyond.--MrFourVideoCards 03:43, 24 July 2011 (PDT)

Start/End Fade Dist/Pixels

I think this needs more detail. It may seem like a simple concept, but what are the optimal numbers here? I know it all depends on the position of the model and the map design, but after experimenting with decompiled EP2 maps, I've noticed that many of the props share different values of Start/End Fade Dist/Pixels. Is there really an optimal system in use for punching these numbers in or is it just by the feel of the level designer? —Mattshu 10:33, 1 May 2012 (PDT)

John Carmack as BSPs inventor

I changed the line which sounded like John is the inventory of the BSP tree when in fact he is not.