Chief Blur Buster wrote:The average lag of these monitors at 240hz is 3,7ms to 4,5ms, well that´s only a 1ms lower than a good and fast 144hz monitor that usually has 4ms to 5ms input lag at 144hz (LG 24GM79, Asus MG248Q, Asus VG248QE for example).
When it comes to real world CS:GO tests, to input lag ranges, MIN/AVG/MAX, the lag difference becomes bigger.
For example, see the 1000fps VSYNC OFF
(of Jorim's GSYNC101 tests) for 60Hz vs 144Hz vs 240Hz.
Currently 1000fps CS:GO VSYNC OFF via high speed camera (true button-to-pixels test)
MIN/AVG/MAX, for 1000fps VSYNC OFF, from 40 passes in CS:GO
60 Hz = 14ms, 21ms, 27ms (13ms lag-randomziation factor)
100 Hz = 14ms, 17ms, 21ms (7ms lag-randomization factor)
120 Hz = 12ms, 16ms, 19ms (7ms lag-randomization factor)
144 Hz = 12ms, 15ms, 18ms (6ms lag-randomization factor)
240 Hz = 12ms, 14ms, 16ms (4ms lag-randomization factor)
Here's an animated gif
that compares lag spreads for 60Hz and 240Hz on the same monitor. The 1000fps VSYNC OFF test is the second last set of 3 bars near the bottom of the graph.
For 144Hz vs 240Hz -- while average
is only ~1ms-1.5ms difference, the lag spread
reduces by twice the average from 4.1ms scanout versus 6.9ms scanout (~2.8ms lag-spread from pure mathematics, ~2ms from 40-passes of real world lag test)
Lag randomization affects aiming to the point of (aiming speed in pixels/sec * lag)
. Mathematically, when doing fast-flick turns at 5000 pixels/second (along the screen surface) while trying to visually time the stop-your-turn-with-crosshairs-on-enemy. 6ms lag-randomness means , 5000 x 0.006 = 30 pixel
range of overshoot/undershoot error. 4ms lag-randomness means 5000 x 0.004 = 20 pixel
range of overshoot/undershoot error. If you turn slower, this is less important, but many eSports players flick faster than 5000 pixels/second to essentially stop their flick to land their crosshairs on an enemy.... Or if you're taking your time to visually aim, say, 3000 pixels/sec, it's till a significant improvement to play on a (good sample) of a 240Hz monitor. Lower lag randomization means less time spent doing back-and-fourth aiming corrections. So even if you can't "feel" the lag directly, it manifests itself as marginally increased aiming errors. In this case, even 480Hz and 1000Hz monitors makes a hell of a lot of sense eventually (once they're better quality than today's 144Hz monitors) since those can reduce lag-randomization to practically nil.
Now that said, if the 240Hz monitor is unexpectedly crappy (adds more blur than it should have!), then yes... but see, above, I've got pretty clear-looking 240Hz vastly superior to 144Hz. Somehow, some 240Hz monitors are much blurrier than others, and one user fixed it in an exchange -- and now we need to figure out if this is a one-off thing, or a brand-specific thing.
Lag-randomness error is bigger for framerates under 1000fps, so this is an extreme test that actually lowers lag spreads as much as possible. In reality, the lag-randomization is 6.9ms for 144Hz and 4.1ms(1/240) for 240Hz.
Lag-randomization is because of random locations of VSYNC OFF tearlines. Standard lag tests don't generally handle cover this! Lag randomization range always shrinks at higher Hz, due to the mathematical frequency between screen refresh opportunities.
However, lag spreads is always roughly one refresh cycle, because sometimes an enemy appears shortly before scanout hits a tearline. Or shortly after scanout hits a tearline. In high speed video a monitor is scanning out from top-to-bottom, and by chance, a tearline might appear right above crosshairs or right below crosshairs. (Or likewise, a specific enemy anywhere on the screen -- tearline locations are uncontrolled). This penalizes you with a random lag of (0 ... refresh cycle length) added above-and-beyond monitor's processing lag.
This is from "first-anything-on-screen-reaction" criteria (high speed camera, peripheral vision) rather than "first-single-point-on-screen-reaction" criteria (Leo Bodnar, photodiode oscilloscope, etc) and VBI stopwatches (cable dongle on screen, or Leo Bodnar device). Sometimes the "first-anything-on-screen" lag-stopwatching criteria is sometimes more important for real-world lag, like playing eSports games and using your peripheral vision to spot enemies, since different parts of screen refreshes sooner than others...
Alas, there are many ways to measure lag
, and they output different lag-spreads and lag-comparision numbers.