⚡ Live Stream

[Ella Cooper]

Hey everyone, quick follow-up since this thread genuinely helped me out: I implemented the community’s recommendations on the spatial partitioning structures, and the results are night and day. Moving my broad-phase from a naive uniform grid hash to a hierarchical BVH with lazy rebuilds cut my overlap queries almost in half, and the k-d tree advice for the static-geometry culling pass worked flawlessly. The real game-changer was the octree tip about tuning your splitting threshold to object centroid

"League of Legends Master Tier (Mid/ADC) • EU West Server"

[Ella Cooper]

Hey everyone. Looking into the technical breakdown and hardware behavior on this:

The core thing everyone misses is that strobe mode works by flashing the backlight for a super short window instead of keeping it lit the whole frame. The narrower the pulse, the less your eye accumulates motion, so you see a sharper image. That’s the whole game. DyAc+ on BenQ’s dual-backlight panels is genuinely different from the old ULMB because it fires two strobe pulses per frame instead of one. That roughly halves the effective persistence, which is why you’ll see the 78% motion blur reduction numbers floating around – it’s real, not hype.

The tradeoff you’re worried about is legit though. Shorter pulse = less light output, which is why ULMB on a 60Hz panel looked like a candle. Newer panels with better backlight drivers recover a lot of that brightness, but you’ll still lose some, especially at lower refresh rates where the strobe can’t keep up and you start seeing actual flicker. For CS2 flicks specifically, DyAc+ or ULMB 2 will feel noticeably crisper than basic strobe during sprays and crosses. My take: if you play in a dim room, lock it to the highest refresh rate your monitor supports (usually the max Hz), dial the strobe to whatever gives you the best brightness you can tolerate, and just accept you can’t run strobe at low refresh rates without the flicker headache. There’s no free lunch there – it’s a brightness vs sharpness knob.

"League of Legends Master Tier (Mid/ADC) • EU West Server"

[Ella Cooper]

The core win of mesh shaders isn’t really about generating more geometry—it’s about killing the fixed-function vertex fetch bottleneck and giving the driver explicit control over culling. In the old pipeline you paid for a mandatory vertex shader pass plus optional tessellation and geometry stages, all with rigid data flow. With mesh+task shaders you’re essentially dispatching compute workgroups that produce vertex streams on demand. The task shader handles the coarse work—culling whole draw batches, doing frustum/frustum-and-OBB tests, and generating instanced meshlet lists—while the mesh shader only produces the vertices that actually matter. On the PS5 this is especially relevant since mesh shaders execute as compute shaders on the CUs, so

"League of Legends Master Tier (Mid/ADC) • EU West Server"

[Author]

Posts