nbohr1more

Another:

The old youtube videos have pretty bad motion artifacts but you can kinda see what's going on:

Yes, but Arcturus misunderstood how the system worked so he found out that the grass demo was rendering wrong unless you set LOD settings to highest ( basically off ). Tels and Arcturus were going to work together to improve the demo but Arcturus got busy with stuff and never returned to it.

HAH! Now r_shadowMapSinglePass 1 is slower than 0 ( disabled ). I'm now over 70FPS in BCD tavern. Gonna verify further.

Hmm. Now I have to evaluate a performance regression. 2.10 gives me over 90FPS in this same scene...

Bikerdude is the main author ( other than grayman ). JackFarmer keeps in contact with him off-channel so he offered to assist with this post before Amadeus or Dragofer were available.

Glad to help 2.11 beta install \ upgrade instructions ( thread ) :

If I'm not mistaken, this mission is meant to coincide with the upcoming 2.11 release so I am guessing players will need to be on 2.11 beta-04 ? ( or newer )

I wonder if we are allocating too much of our precision to smaller scales? If we figure out the smallest asset size we might want to account for in polygons ( rings, coins, silverware ) then make anything more than 1.5x smaller merge polys with surrounding geometry or disappear, we might gain another 3 or 4 bits of precision at the larger scales to grow those larger maps? This would might also solve things like invisible geometry slivers causing map bugs or invisible hair-line leaks. Easier said than done though...

Hmmm... "Grid size is 131,072 x 131,072 units x 131,072 (inches) or about 2 x 2 x 2 miles" https://wiki.thedarkmod.com/index.php?title=Limits,_Max,_Min,_Stats,_etc This wiki is a little out-of-date since @stgatilov removed the entity limit in 2.10 ...

Please make sure that these issues only happen in 2.11. If the same things happen in 2.10 or earlier and there are no open trackers then please open new ones.

Got a used RTX 2060 6GB for Christmas! It's pretty strange being able to run TDM at full renderscale 1080p, shadow maps at 1440 size, soft shadows quality 25, 4xMSAA, 16xAF (SSAO, Bloom, and Fresnel mod) and still have gas in the tank for 100FPS or more. I found that Penny Dreadful 3 courtyard performance can be conquered by a better GPU and so can the opening area of Briarwoord Manor ( notorious outliers ) and the finale area of Written in Stone. However, the tavern in BCD is still stubbornly stuck at 40FPS and Scroll of Remembrance is stuck at 37 FPS. I went back to do a performance comparison between 1.07 vs 2.11 beta and found that even with my new GPU 1.07 runs the original Return to the City at sub 30FPS in the opening area. In 2.11 with equivalent settings I get 95FPS with the same map. I was also saddened to see the Serengrove map running under Doom 3 Sikkmod still runs sub-20 FPS. On a lark I tried extracting it to RBDoom3BFG and it was locked at 60FPS with all fancy features enabled. There has been a longstanding stance that Doom 3's CPU usage was a bit heavy but the biggest consumer was GPU fillrate. Based on my testing, throwing fillrate at the problem doesn't seem to be the magic bullet. Given how much better CPU optimization is for both TDM 2.11 and RBDoom3BFG I would guess that a smaller GPU upgrade and a better CPU would have been a better choice. I'm still overall happy with my upgrade and the i7-3770S is still holding it's own pretty well too.

Now I recall where this happened. This is a known artifact for the "Frob Outline" feature. Nuking your config would cure it but you can also just disable that feature.

Merry Christmas !!!

Because no entrants made it this year, all eligible FM authors will be visited by the ghosts of darkmod past, present, and future. The spookiest outcome of all! ( "Ghost of darkmod past" politely asking you to purchase a copy of Doom 3 so he can haunt you... )

I am doubtful this is a legitimate discussion. While AR hud features might end up in TDM VR, no Thief fan is going to use a buzzword laden article about AR business strategies to broach this type of topic. I will suggest we have a Spammer. That said, if this is a subtle Spam campaign to get TDM to add AR so that in-game adverts are shown I say bravo to whatever Spam AI engineer designed this bot. I am locking this discussion. If any legitimate discussion about AR is to had it will be in the thread mentioned by @freyk

This can happen if you have r_useEntityScissors enabled. Please confirm that it happens when that is set to 0.

Is this still happening in the latest SVN ? If so, more GLSL debugging should be visible with r_debugGLSL now as I recall.

Yep. Not sure what was happening but a new compile and compared to 2.11 beta 2 shows matching performance. No issues here anymore. Performance is great. Side note, when I want to get fancy with compiles, I add a flag for my own CPU arch: cmake -DCMAKE_BUILD_TYPE="Release" .. -DGAME_DIR=/home/user/darkmod -DCMAKE_C_COMPILER=gcc-11 -DCMAKE_CXX_COMPILER=g++-11 -DCMAKE_CXX_FLAGS=-march=ivybridge https://gcc.gnu.org/onlinedocs/gcc/x86-Options.html ( doubtful that much is gained at all )

No worries. I experienced a noticeable performance downgrade ( not as extreme ) so I will test again this evening and get hard numbers if anything is off.

I am sure that the glprogs folder takes precedence but if you want another point of confirmation please try unpacking the tdm_base01.pk4 then overwrite the glprogs folder inside it and repack to see whether things improve.

For the 10211 results, did you restart TDM between setting changes? What happens with r_shadows 0 in 10211 ?

I believe @stgatilov was asking that you provide comparisons of the basics: 1) Hard Stencil beta vs svn 2) Hard Maps beta vs svn 3) Soft Stencil beta vs svn 4) Soft Maps beta vs svn That should triangulate the affected area. On review of the code change, in theory it "should" perform much better since the blur operations are done on the GPU (thus freeing the CPU for other work) but I guess there may be a pathological GPU pattern that looks good on paper but the hardware or compiler doesn't execute well.

Nvidia here... I lose about 5 to 15% performance with this change. Since the problem is with AMD on Windows, perhaps a shader branch for Linux in general ( AMD on Linux supposedly gets worse with this change too ) ? Not too catastrophic but is gonna be a little tough to explain in a 2.11 release article. It would be cool to know what happens on Nvidia + Windows though...

Rant: