Sharing my newly released paper on optimization testing in Source engine

Quote from El_Exodus

I'm writing a guide aswell which is more or less covering the same thing. "The dos and don'ts of Counter-Strike Level Design" I guess i am late now. :< Although i write about a few more topics. But yours is offering a better reading experience I hope i can finish it soon.

The topic is vast and there is always room for more articles; the more the merrier

Quote from nikkoship

Yikes, I need to read alot of these. I've read a bit of Man Vs. Engine and it was super helpful. It's awesome that there is someone like you out there to help with the really technical side of map-making in source.

Thank you just trying to help with whatever I can.

Quote from laminutederire

@will2k Are you afraid source 2 will change everything?

Not at all

change is needed, rather required; today, not even tomorrow.

Source is a nice engine but it's old...read limiting and frustrating at times. What worked quite well for Carmack & Co in 1995 does not necessarily work well in 2015. Software rendering is no more and hardware limitations are more or less a thing from the past.

Granted that Source is very easy to get into and its additive geometry is easy to work with, but once you get more acquainted with it, you start to realize the huge constraints: brushes, invalid structures, concave shapes, rounding errors, bsp, pvs, baked lighting...

From the little I read about source2 over a year ago, I know that brushes are replaced with meshes, concave shapes are now allowed and bsp is no more, which leads me to speculate that it will be a real time GPU-dependent occlusion for the visibility calculations.

Source is a solid engine, don't get me wrong but now is the correct time to make the move otherwise UE4 and Unity are gaining ground rapidly and will start taking away big chunks of Source's market share and "engine of preference" among designers/modders.

Yes, this will mean that whatever I know about Source might become obsolete with source2 and I'll have to learn everything from scratch about the "new" technology but that's fine; sitting for too long in the comfort zone makes one complacent :). A change is necessary to shake things up and open new horizons.

And in case I don't like what I see in source 2 then there is always UE4

Quote from laminutederire

I wish they will hurry

I'm no expert, so correct me if I'm wrong, but... aren't the pbr pipelines very powerful and another argument why artists will move to something else?

Who says Source2 won't have PBR somewhere down the line or a similar rendering technique for realistic lighting calculation

Source2 is not "officially" released yet (as in available for free to the public in its complete form as revealed earlier this year); the portion that got into Dota2 couple of months ago is the tip of the iceberg as far as I can tell. Many features/tools are still unreleased and Valve being Valve, most of the technical features are still not fully disclosed.

I guess we will have to wait for a major release from Valve (HL3?? or L4D3) to push the engine to its limits and showcase its full technical features.

I'm aware that UE4/Unity/Frostbite engines already have PBR with realistic/stunning results and streamlined workflow for artists/programmers, so that's an advantage over source 2 right now, but remember that source1 was an "incremental" engine, benefiting from incremental new features along its existence (ep1, orange box, l4d, portal2, etc.). I believe source 2 will follow a similar path, and if PBR is not there at release, it will find its way somewhere in a future update.

I'm not defending source 2 though (haven't used it nor do I know its technical features inside-out); even with PBR and other features implemented, it could turn out to be an "inferior" engine compared to the current ones that incorporate PBR. In this case, there could be an "exodus" towards UE4 or Unity...time will tell