changed the way i cull subchunks courtesy of @bnut. i extract the world space frustum planes from my view-projection matrix, normalize, then do `dot(plane.normal, subchunkCenter) + plane.dist` and check against the bounding sphere radius (0.5*sqrt3*side)

occlusion culling now runs in 0.43x the time but also the culling is up from 43% to 74%, which means the game runs much faster. at 12 chunk distance fps above ground is up from 60 to 110!

interestingly, if i switch from 4 subchunks per axis (64 per chunk) to only 2 (8 per chunk) the entire occlusion culling pass finishes in just 7ms even on my underpowered workstation at 12 visible chunk radius

unfortunately the culling% does go down from 71% to 57% and so fps goes down from 110 to 90, but that might be worth the trade off 🤔

@eniko These probably count as optimizing more than needed, but a few suggestions:

1. test without the early out and see if it's faster. It's more ALU work, but because of how deeply pipelined CPUs are you really want to keep the code flow in small loops very predictable.
2. assuming you're targeting 64-bit platforms, you should have access to SSE2, which would let you test 4 chunks against the view frustum at once. Depending on how the memory latency works out, that can net another 2-4x.

even if i set the visible chunk radius to 16 (which is quite far!) the occlusion culling runs in 11ms* and my fps above ground is 55

which i would say is solid for this early in the project

*with 8 subchunks per chunk

Oh. Most of the gains are from the frustum culling. I can cull a significant % of chunks if I use 64 subchunks per chunk, but that is also very very slow. But it does bring cull % up to like 75%

But just frustum culling chunks culls like 45-50% of chunks and is very very fast :|

8 subchunks per chunks is just a regular amount of slow but also not very effective

So hm. Well I guess at least I learned how to do efficient frustum culling? Womp womp 😔

oh hey i got it down to 54ms from 93 at 16 visible chunk radius with 64 subchunks per chunk by only processing subchunks that have visible faces

i was using visible *blocks* before. but a theoretically visible block (not void/air) surrounded by blocks has no visible faces, so can't actually be seen

oops i had a bug that was causing a bunch of subchunks to not get occluded when they should be. so now the number of depth occluded subchunks is up by 50% and there's 1 depth occluded subchunk for every 2 frustum culled subchunks

Had an idea to try tomorrow

Only render occluders nearby the camera and skip depth culling, because far away occluders will be tiny and unlikely to occlude much

Then depth cull far away (sub)chunks, since they're tiny and so only cover a few pixels and so are more likely to be culled

Actually it could be interesting to treat subchunks like mipmaps 🤔 so level 0 is 1 block per subchunk, level 1 is 2, 2 is 4, 3 is 8 and 4 is the whole chunk

Then I can use lower levels up close for better occlusion and higher levels far away for quicker culling

aha! i might be onto something! only rendering occluders into the depth buffer for nearby subchunks and only depth culling far away subchunks dropped the run time from 0.21x original to 0.14x original

i bet if i reduce the number of subchunks far away by making subchunks larger i can get even better gains

so uh, wow, holy shit!

my new subchunk levels method (where each level is smaller) is a 7x improvement over my last attempt, and my software occluder now runs in 0.02x the time it did before i started optimizing. in other words, it runs 50x faster now! for 12 visible chunk radius its down from 155ms to fucking *4*

interestingly though, even though i have 3 levels, 16, 8 and 4 blocks, i didn't even need the middle level. i just switch straight to whole chunks for depth occlusion

in hindsight this makes sense. occluders should be small, since they set up the depth buffer, but then occlusion culling can be done on whole chunks since it doesn't matter how big the area you're testing at one time is so long as the occlusion map is filled up properly

so next i might try rendering single blocks into the occlusion map and depth culling entire chunks always and see how that performs

16 visible chunk radius now runs at 68 fps above ground (1 short of being truly nice 😔) and the occlusion culler runs in only 5ms

simplified everything and split occluders from depth culling candidate chunks

also trying to be more picky and granular about what occluders to use, so now i'm rendering single blocks nearby the camera. hopefully more granular occluders will improve the culling percentage

you can call me weird but it is very fun to build a little box around myself and watch my fps go up from 110 to 450 because of the depth occlusion system

hmm hmm so my subchunk levels go 1³, 2³, 4³, 8³, 16³

but im betting anything above 2³ is gonna be exceedingly rare since a new requirement is having a block with an exposed face. though i havent tested this

but i feel like maybe 1³, then 2²x1 4²x1 8²x1 16²x1 for each axis would probably work a *lot* better given how human construction tends to go

and also ground planes/cliff faces

i'm *sure* you'll all find this as riveting as i do, but while there are loads of 8x8x8 areas of fully opaque blocks with at least one visible exterior face, there appear to be literally no 16x16x16 areas of the same description in any given chunk >_>

here's photographic evidence of all the 8x8x8 blobs out in their natural habitat

so i've tweaked all the parameters and above ground its now culling 75% ish of chunks that survive frustum culling. even at 24 chunk radius it runs in ~8ms so about half a frame (and it's threaded so that's fine)

and i'm getting 75 fps above ground at 16 render dist, and 115 at 12. and indoors or in caves its more like 300-400

so i'd call this feature a success and also done for the time being!

being in a cave make fps go zoom! :3

(lol the alt badge hides the fps counter. it reads 450)