Blur Busters Forums

So I finally decided to stop being lazy and test my new BenQ XL2420G's input lag. I conducted this test in classic mode, classic blur reduction, and g-sync mode. Please keep in mind that these results are input lag and response time, so the results will be higher than other testing methods. The results below are quite interesting and I'm curious if you guys can explain it to me.

Here is my setup:
-Leo Bodnar Lag tester, 1080P @ 60hz
-Tested classic and blur reduction mode using the HDMI input
-Tested the g-sync mode using the displayport input, had to use an active hdmi to displayport adapter to make it work.

Classic mode:
Top: ~2.3ms
Middle: ~10.0ms
Bottom: ~17.3ms

Classic, BenQ Blur Reduction:
Top: ~6.8ms
Middle: ~15.0ms
Bottom: ~23.4ms

G-Sync mode:
Top: ~11.1ms
Middle: ~14.2ms
Bottom: ~18.1ms

If we compare results, obviously classic has the best time with 10ms in the middle, G-sync lags slightly behind at 14.2ms, and the blur reduction is slightly more at 15ms. My main concern was the g-sync module in the monitor may be adding some lag compared to the classic mode. This may be true, but the adapter may be adding lag as well. If we look at the difference between the top and bottom of the screen the results are quite interesting.

Difference between bottom and top of the screen:
Classic: 15ms (which is the expected 15ms refresh top to bottom of most LCD panels).
Blur Reduction: 16.6ms (highest out of the three modes)
G-Sync: 7ms (half of what I would expect...interesting)

So why is the G-Sync taking less than half the time compared to the other two modes? I was curious, so I re-read the displaylag.com review (http://www.displaylag.com/benq-xl2420g- ... or-review/) and found that they had very similar results for the different modes.

So I guess I have two questions:
1. Why does g-sync have less of a change from top to bottom of the screen?
2. Why does g-sync mode add 4ms of input lag? I guess this could be the adapter or just the nvidia module being slightly laggier than classic mode.


A few things people may find interesting:
-Both HDMI inputs in classic mode had the exact same lag, don't worry about which input you use.
-I was unable to capture ULMB input lag because my adapter stopped working for some reason. I've requested a replacement and will update this thread if I am able to get it to work.
-My LG W2363D had 11ms of input lag
-My Panasonic P55GT50 had 56ms of input lag (although this is probably inaccurate, apparently the Leo Bodnar tester does not work well with plasma screens).
-Again, all these results were captured at 60hz. So the absolute minimum you can get is around 8ms. If I was able to test at 144hz the results would be slightly faster.

gsync scans out as fast as 144hz which means the difference between top and bottom is ~7ms

it's probably the adapter though

flood wrote:gsync scans out as fast as 144hz which means the difference between top and bottom is ~7ms

it's probably the adapter though

Hmm... Well the Leo Bodnar lag tester outputs at 1080p and 60hz, so that would be very odd.. unless the adapter is changing the signal from 60 to 144hz. The numbers do seem to match up...

Thank you!

wait a second are you saying you got gsync working on a source that isn't an nvidia card?

on my gsync-modded vg248qe, only nvidia cards would allow gsync mode on the monitor

flood wrote:wait a second are you saying you got gsync working on a source that isn't an nvidia card?

on my gsync-modded vg248qe, only nvidia cards would allow gsync mode on the monitor

The lag tester connects directly to the screen and doesn't go through the video card at all. It's going through the g-sync displayport but g-sync isn't actually enabled.

i see, but the 7ms difference means that the scanout speed is that of the normal 144hz mode

which makes sense if gsync is running...

this is weird

flood wrote:i see, but the 7ms difference means that the scanout speed is that of the normal 144hz mode

which makes sense if gsync is running...

this is weird

I agree and that is why I created this thread. I have no other way to measure input lag, so I was curious if other methodologies showed different results. Ideally the g-sync module has the exact same input lag as the BenQ classic module. However, all data I have seen has shown the opposite, but all that data has been collected using the Leo Bodnar device.