TestUFO tests failing to sync ONLY with multiple monitors

NEW for 2017: Discussion about the testufo.com Blur Busters Motion Tests. Widely used by enthusiasts, display tweakers, YouTubers reviewers, monitor manufacturers and VR headset makers!
kooz
Posts: 9
Joined: 01 Feb 2017, 15:48

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by kooz » 07 Feb 2017, 20:34

All three are using DisplayPort cables. I restarted the graphics driver using CRU several times, and every time I ended up getting a different combination of refresh rates at each monitor.

For example, staring with the table I posted earlier, it showed:
1.) 119.2Hz 2.) 120.4Hz 3.) 120.4Hz
-restart graphics driver-
1.) 120.4Hz 2.) 119.2Hz 3.) 119.2Hz
-restart graphics driver-
1.) 119.2Hz 2.) 119.2Hz 3.) 120.4Hz

This occurred without me physically touching anything except my mouse. It appears to be changing without any rhyme or reason.

I'm wondering if the "current frequency" value given by MonInfo is a nearest approximation between 120.4 Hz and 119.2 Hz and its just changing what value it rounds to due to minor clock drifts or something.

Could be a red herring.

I ended up switching back to "normal" triple-monitor mode today, as it was nearly impossible to be productive with Windows treating all 3 as one gigantic wide display... The longer I've been trying to fix the issue, the more noticeable these little hiccups have become.

I'll try throwing a dual-link DVI cable into the mix tomorrow for the hell of it.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by Chief Blur Buster » 08 Feb 2017, 12:47

kooz,

That can't be too fun, trying to get three 120Hz monitors to perfectly sync to each other.
I wonder if adding a 2nd card to the mix helps or hurts things.

I wonder if this is something that is worth reporting to NVIDIA.

It is wholly possible that this is an application dotclock calculation error, but perhaps use www.vsynctester.com (one window, Chrome, maximized only on specific monitor, close all other windows). Wait for 30 seconds and write down refresh rate. Currently, if you only have one Chrome window, it tends to like syncing to the monitor that it's currently maximized on. It's able to measure approximate fractional refresh rate (while TestUFO intentionally rounds it off).
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

kooz
Posts: 9
Joined: 01 Feb 2017, 15:48

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by kooz » 08 Feb 2017, 17:34

Chief,

First off, thank you for your continued assistance! I had no idea that vsynctester.com existed - it is excellent at providing a still-image visual representation of the issue I'm having.

I took a screenshot after it stabilized (as much as it was going to get) for each of my 3 displays. The hiccup spike is occurring once every other pass, each time appearing slightly further to the right of where the previous spike occurred on the graph. So, without a doubt, it is happening at a consistent interval.

Image
Image
Image

I'm not sure if the calculated refresh rate is being averaged out or not, but if so, there is a maximum of 0.000385 Hz difference between the displays, according to my screen grabs. I don't know the math offhand to figure how often this disparity would add up to a full frame out-of-sync, but I doubt it calculates to the ~2-3 second frequency I'm experiencing.

Also, the high refresh rate, while it would seem to be a contributing factor, seems to be irrelevant. 120 Hz is what I'd prefer to use on a daily basis, but the issue is present with all of the refresh rates I've tried (144, 120, 100, 85, 60).

I would love to drop $350 on another card to test its effects, but I can't quite afford to do that at the moment, haha... :roll:

Contacting Nvidia may indeed be worth a shot.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by Chief Blur Buster » 11 Feb 2017, 17:34

Thanks for posting these results, kooz!
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

kooz
Posts: 9
Joined: 01 Feb 2017, 15:48

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by kooz » 13 Feb 2017, 23:02

I'm beginning to think this is less of a hardware/driver-level issue, and more OS/software related...

I came across a few recommendations for a program called DisplayFusion after looking through some threads in the Nvidia Surround forums and decided to give it a try. Essentially, the idea is to leave the graphics driver in Surround mode and have the software divide up the single display into user-defined segments. This should work fine, since I can run all 3 monitors completely stutter-free while running in Surround mode.

I got it set up and working on a basic level, but came to a screeching halt shortly thereafter when I started noticing frame hiccups while DisplayFusion was running, despite the driver running in Surround mode.

It was extremely easy to verify that the new software was the culprit in this case - I pulled up vsynctester.com in Chrome and verified the hiccups were occurring, then I simply right-clicked and exited DisplayFusion via the system tray icon, at which point the test immediately went back to running stable with no spikes. I didn't need to reload the browser or the page or anything.

This leads me to consider the possibility that there is some inherent issue with the way Windows (10 Professional x64) manages multiple desktop "spaces". Or maybe DisplayFusion just happens to have some overhead that is producing a very similar yet entirely different issue... It seems unlikely.

I think I'm going to spin up a temporary drive with Windows 7 to see if that makes any difference.


UPDATE:
I decided to do some further testing with DisplayFusion since the result just wasn't making sense to me. It turns out, the only reason it was stuttering was because I was in the Monitor Configuration dialog in the program. I assume the stutter was occurring because it is polling hardware during that time! :lol:

So, my issue is essentially resolved... It's a 3rd party software fix, but it works nevertheless.

Nvidia Surround + DisplayFusion = win

Check it out - glorious! :D
Image

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by Chief Blur Buster » 14 Feb 2017, 22:28

Fantastic, I am glad you finally came to a compromise on this!
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

kooz
Posts: 9
Joined: 01 Feb 2017, 15:48

Re: TestUFO tests failing to sync ONLY with multiple monitor

Post by kooz » 15 Feb 2017, 00:32

Thanks for following my story along the way. :)

Since my previous post, I have encountered a few quirks/annoyances that keep DisplayFusion from being a truly seamless substitute for native multi-display mode; mostly related to the way windowed apps behave when instructed to go full screen (i.e. media players, browsers, etc). They are not on top of the taskbar, nor are they being constrained to a single display.

You guys don't need to hear about that, though. I feel like this is getting off-topic. I'm sure I'll find a way to overcome these minor issues.

I'll stay subscribed to this thread though. I would like to know if anyone else is encountering the same issue, and would especially like to see a solution that doesn't involve a 3rd party software workaround.

Post Reply