Input lag @ non native resolutions

[Mr1991]

Hi, I'm looking to upgrade from a 144hz display to a 240hz, I've been browsing various sites/looking at reviews and I've mostly boiled it down to one of these:

HP Omen x 25f
Acer Nitro xf252q

The only problem I have is I can't find any info on their input lag, specifically at non-native resolutions, if they have an increase in delay on different resolutions, they're not all that usefull to me, this is rtings review of the zowie xl2546: http://puu.sh/EIaee/11ef707003.png as u can see they've listed the input lag on different resolutions as the same, but their other 240hz monitors reviews they haven't for some reason: http://puu.sh/EIalf/2b7bd24327.png

Wondering if anyone can give me any insight to this, before I make a purchase, thanks !

[Chief Blur Buster]

It depends --

Technically it's possible to do scaling virtually laglessly (line buffered scaling -- less than 10 microseconds), but it has historically been extremely difficult to test for.

For monitors that do not do scaling, you have to do GPU scaling. Fast graphics cards like GeForce RTX series can do GPU scaling in just tens or hundreds of microseconds, which is actually less than the input lag difference of an LCD at 60F versus LCD at 70F (temperature-induced latency, where monitors responds slower in colder temperatures). So why care about GPU scaling when it's less input latency than room temperature difference of summer versus winter? Forgetting a smartphone in a car in the winter, or a frozen LCD, you see them respond really slow. But what many don't know that even just a few degrees temp difference can create millisecond or sub-millisecond-scale latency differences.

Certainly, it is possible it is important at the millisecond timescales (there are crappy scaling latency differences -- some monitors have historically done a bad job of scaling), but you can bypass laggy monitor scaling with an ultrafast GPU that does GPU scaling darn nearly laglessly anyway. A full 1080p framebuffer scaling is benchmarked at approximately ~80 microseconds (0.00008 second) on my GeForce GTX 1080 Ti Extreme at its maximum rate, and that was a scaling test in C# (a slower programming language than C++). Even a 1-2 degrees room temperature difference creates more input lag difference than modern optimized GPU scaling nowadays.

Now, there are good reasons to stick to monitor-side scaling (e.g. consoles and other limitations), and that's where you have to pay attention to benchmarks. But for PC, it's actually a concern gradually falling below noisefloors. Even I can see 0.25ms vs 0.5ms MPRT differences with my unaided human eye (TestUFO, Photo, Toronto Map Test, 3840 pix/sec), but I can't tell a 0.25ms difference in latency.

[Mr1991]

So I assume Rtings would be using display scaling and not gpu, as that would defeat the purpose of the test, the reason I ask is because my LG 144hz monitor has terrible input lag when scaled down from 1920 x 1080, to 1280 x 720, and I guess my gpu isn’t fast enough because gpu scaling definitely feels even worse.

If the difference in the resolution scaling on the newer 240hz monitors was 1millisecond or less, (or microseconds as u say) that wouldn’t be a problem at all, I just don’t want a 10ms increase, as it feels with my current monitor. So would it be safe to assume that most of the high end 240hz monitors would have minimal difference? (If the Zowie can do it w/o penalty)

[Chief Blur Buster]

So I assume Rtings would be using display scaling and not gpu, as that would defeat the purpose of the test, the reason I ask is because my LG 144hz monitor has terrible input lag when scaled down from 1920 x 1080, to 1280 x 720, and I guess my gpu isn’t fast enough because gpu scaling definitely feels even worse.

If the difference in the resolution scaling on the newer 240hz monitors was 1millisecond or less, (or microseconds as u say) that wouldn’t be a problem at all, I just don’t want a 10ms increase, as it feels with my current monitor. So would it be safe to assume that most of the high end 240hz monitors would have minimal difference? (If the Zowie can do it w/o penalty)

It's a very good question to ask RTINGS, and see what their lag testing methodology is for GPU scaling modes.

[Mr1991]

Another quick question, as I’ve seen u talk about them, would you say the new 240hz ips displays will have similar low input lag as the 240hz tn panels, and do u think they’re worth waiting for?

[skylit]

I tried a 27" 1080p 240hz ips. I personally think the latency is similar, Slightly "different" than Gsync module panels.

My overall issue is that the response time looks off and is noticeably visually "laggier" than the original 240hz TN's. I also suffer from "IPS glow" which I didn't know was a real concern.. but it bothers my eyes and makes gaming annoying.. 1080P 27" can also be a nusicance just in terms of PPI.. Everything is a lot more blury at my normal distance. Not worth the trade off subjectively.