Page 4 of 6

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 26 Apr 2017, 08:06
by Dew
I also tried RealNC's trick by adding a custom refresh rate with CRU and capping with the in-game limiter. Unfortunately I still feel input lag in Overwatch even after being very generous with the gap in refresh rates. I also noticed frametimes were very consistently spiking up and then slowly rolling down until it spiked up again creating a smooth wavy frametime graph in Afterburner. I have a feeling the reason I'm noticing input lag is because I'm using an AMD card and their pre-rendered frames works differently with little explanation or instructions on how to tweak it.

I created the FlipQueueSize registry command and set it to 3100 for a value of 1. I changed the Main3D key to 3100 for a value of 1 and I changed the Main3D_DEF to 1. Main3D_SET was also changed to 3100 for a value of 1.

I have no idea of this properly sets pre-rendered frames to 1. These registry keys and their claimed functions were found in RadeonMod. It sure is easier to change pre-rendered frames on Nvidia cards....

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 26 Apr 2017, 11:36
by Chief Blur Buster
Sounds like AMD has very different lag-creating framebuffer backpressure than NVIDIA in this case.

It'd be interesting if we could figure out how to fix this... (I imagine it should be fixable.... but how?)

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 26 Apr 2017, 12:37
by RealNC
straussmanover wrote:Sidenote with CRU: For the sake of messing around, I tried out a few more "aggressive" changes to the refresh setting, 115, 118, 122, 125. So after these changes and a driver restart (even a machine restart), I pulled up some UFO tests, and it was always reading 120hz, no matter what I had changed the refresh to, and ULMB was undoubtedly still active. So, this has me a bit concerned that the change from 119.990 to 120.010 was having no impact. (Perhaps the monitor or driver has some sort of "deviation limit?" Not allowing the rate to be changed too far from its intended speed?)
Just use:

https://www.vsynctester.com

to detect your exact refresh rate. When you test without ULMB, and then switch ULMB on, is there a difference in what vsynctester reports?

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 26 Apr 2017, 13:30
by Chief Blur Buster
Although TestUFO will tend to roundoff 119-121Hz to 120Hz --

TestUFO should still display 115 and 125.

I'll be creating a new TestUFO "detect refresh rate" pattern, for ultraprecise fractinal refresh rate detection.

I love vsynctester -- and the creator (duckware) collaborates with us on browser VSYNC issues. But I certainly shouldn't send all my traffic to vsynctester, once I create that "TestUFO Detect Refresh Rate Test", shall we....? ;)

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 26 Apr 2017, 13:50
by RealNC
The more the merrier ;)

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 01 May 2017, 08:16
by straussmanover
RealNC wrote: Just use:

https://www.vsynctester.com

to detect your exact refresh rate. When you test without ULMB, and then switch ULMB on, is there a difference in what vsynctester reports?
Tested with the utility, looks like my display runs about 0.005hz slower than the CRU setting, so I just bumped it up to 120.015 to get it to run at 120.010. No change between ULMB on and off.

Everything is squared away now it looks like. I really appreciate the help!

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 18 May 2017, 04:46
by drmcninja
Previously I had 144Hz/154fps (in-game limiter) and it felt good.

I'm experimenting with keeping 120Hz Lightboost on and keeping FPS uncapped since my computer will fluctuate in FPS from high 100s to low 200s.

I've read on here that Lightboost mode definitely increases input lag by ~4ms (can someone re-confirm that?), so I figure uncapping the FPS and letting it go higher will compensate for the additional input lag somewhat.

In other words:

120Hz Lightboost = +4ms
Uncapping FPS = -2ms (on average)

Is my thinking correct here?

Also, does Lightboost-induced input lag get worse with brightness set to max?

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 18 May 2017, 06:42
by Sparky
I've read on here that Lightboost mode definitely increases input lag by ~4ms (can someone re-confirm that?), so I figure uncapping the FPS and letting it go higher will compensate for the additional input lag somewhat.
The cause of that is that it takes some time to scan out a frame to the LCD. If the backlight is on continuously, you see the picture change as the LCD is being updated, while with lightboost, you only see it when the backlight pulses(after the whole screen is updated). That 4ms is likely an average figure, the difference in delay would be higher at the top of the screen and lower at the bottom of the screen.

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 18 May 2017, 12:46
by KevOW
Dew wrote:I also tried RealNC's trick by adding a custom refresh rate with CRU and capping with the in-game limiter. Unfortunately I still feel input lag in Overwatch even after being very generous with the gap in refresh rates. I also noticed frametimes were very consistently spiking up and then slowly rolling down until it spiked up again creating a smooth wavy frametime graph in Afterburner. I have a feeling the reason I'm noticing input lag is because I'm using an AMD card and their pre-rendered frames works differently with little explanation or instructions on how to tweak it.

I created the FlipQueueSize registry command and set it to 3100 for a value of 1. I changed the Main3D key to 3100 for a value of 1 and I changed the Main3D_DEF to 1. Main3D_SET was also changed to 3100 for a value of 1.

I have no idea of this properly sets pre-rendered frames to 1. These registry keys and their claimed functions were found in RadeonMod. It sure is easier to change pre-rendered frames on Nvidia cards....
It's already set to 1 by default

Re: Overwatch -- Lightboost 120Hz + Fast Sync

Posted: 28 May 2017, 06:06
by drmcninja
jorimt wrote:
KevOW wrote:Thank you Real and Chief for the quick response. I'm curious if any of you ran tests similar to Battle(non)sense

https://www.youtube.com/watch?v=sITJ3V_fyv4
RealNC wrote:We did G-Sync input lag tests for CS:GO. Nothing for fastsync at this point though, sorry.
No, I actually did do a Fast Sync comparison in my original input latency tests, albeit only at 144 Hz:

Image

It didn't add much latency at 144 Hz uncapped (fps_max 300 in CS:GO); this will of course vary depending on the current max refresh rate/sustained fps ratio. The last official word from Nvidia was that Fast Sync has around 8-15ms more latency than v-sync off on average:
https://www.youtube.com/watch?v=xtely2G ... t=1h34m21s

For the upcoming part #3 of my ongoing G-SYNC 101 article however, I will be comparing G-SYNC input latency against v-sync off, v-sync on, and Fast Sync across all available refresh rates, with multiple scenarios covering uncapped and varying levels of capped framerates.

Keep an eye out at the top of the below thread, along with the Blur Busters home page and social media pages for updates:
http://forums.blurbusters.com/viewtopic.php?f=5&t=3073
I'm trying to enable Fast Sync, but I'm not sure it's working. How do you enable it?

I enabled it for Overwatch.exe and Overwatch Launcher both in Nvidia Control Panel, but in-game my FPS is still showing 300. When I lowered it to 140fps with the in-game limiter, I could tell my FPS just dropped significantly, everything began to stutter and mouse movement felt different. My refresh rate is 120 so if Fast Sync was really enabled, it should've been outputting 120fps the entire time so I shouldn't have felt a difference no matter what the fps limiter was set to, no?