RLCScontender wrote: ↑16 May 2020, 13:23
axaro, it's actually closer to +8 to 10ms(minus the small input lag reduction of going up +40hz) input lag because adaptive sync AND backlight strobing is on at the same time.
i had to use bluetooth(insteaD OF WIRED) on my ps4 controller to avoid a big latency hit if i used ELMB-SYNC. If however you don't notice much changes in input lag, more power to ya. But for me, i do notice it and a dexterity and timing heavy game like rocket league, that 8-10ms added input lag will determine if i get a 65 MPH power shot on target or a 22mph soft hit resulting in the opponent getting a free goal on my falling recovery.
if however u aren't able to notice it, more power to ya. For me, i do notice it unfortunately since the game i play requires a lot of precision and dexterity.
+8/10ms? where are you getting these numbers from? Am I missing something?
Enabling VRR generally cause a +0.1/0.2ms increase in input lag (probably due to increased signal processing), capping the framerate is actually the best way to reduce input lag unless you can consistently output 280+fps.
If you are considering a ps4 which has a 60hz output we both know that low hz input lag is higher, but I didn't buy a 280hz to plug a console (and I don't see the point of 60hz strobing if the OD is stuck at 80), couldn't you notice the overshoot at 60hz from ELMB-Sync? It should be unusable sub 144hz (I tested it on Overwatch at 60hz and it's bad, it starts to be good at 180+fps, the sweet spot is 200-280fps, with just 1.8% overshoot at 200hz with OD80 if the same tests from the VG279QM apply to the 24.5")
Sorry but I can't notice a speculated "8/10ms" difference, in theory the penalty should be VRR penalty + Strobing at the beginning of each refresh, which translates to 0.1/0.2 + 1.1 = +1.2/1.3ms with Asus' ELMB(not sync)
With ELMB-Sync's devil horn implementation the strobe duration is a little bit longer, 2.5ms at 240hz and 2ms at 280hz(If my calculations about ELMB-Sync are correct there's a 56% duty cycle at 280hz and 60% at 240hz)
This means two things:
- The strobing cycle is 0.9ms longer at 280hz and +1.1ms longer at 240hz than Asus's standard strobing (Elmb duty cycle is basically 1⁄4 of a refresh, ELMB-Sync is around ½ of a refresh)
- The longer duty cycle further eliminates any form of crosstalk, I honestly can't see any crosstalk even with a dark background around the moving frogs/Ufos.
So the question when using ELMB-Sync is: would you trade half of a 4.16ms refresh for perfect strobing?
I don't notice any crosstalk, even my 480fps camera test can't detect any crosstalk (
https://www.youtube.com/watch?v=H1uvrRqvJHs), fluctuating fps are imperceptible during strobing and there aren't any microstutterings (I followed Chief's tips on how to prevent any form of microstuttering during ELMB)
You can argue that the "devil's horn" pulse method is inefficient, but there isn't a single game where I would disable ELMB-Sync as long as I can consistently output 180+ fps.
Edit: I just read "Strobe Crosstalk: Blur Reduction Double-Images", it confirms that
Strobe duty cycle (flash length) also plays a role., confirming my hypotesis about Elmb-Sync having less crosstalk than ELMB due to an longer strobe duty cycle.
) so I'm stuck playing r6s at 240hz, I'm thinking about tweaking the config file and set it to "Read Only" since it reverts to 240hz after selecting 280hz.