ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Many sites including LinusTechTips, RTINGS, TomsHardware, and others use the free Blur Busters pursuit camera invention. Now also available as a rail-less smartphone wave, too!
Post Reply
User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Post by Discorz » 08 Apr 2020, 10:27

Images below compare how differently AW2518HF (freesync monitor)(and probably on some other monitors too) behaves on AMD vs NVIDIA GPUs.
Many people on Internet reported they have corona/white trail artifact with AW2518HF but my experience was very different and better. I figured researching this might be helpful for potential unsecure G-SYNC Compatible buyers. So I did it and this is how you can fix it by using settings from best result ufos:

AMD GPU: (tested myself with RX 580 and RX 460)
Over drive modes didn't do anything for me. "Normal", "Fast" and "Super fast" modes looked the same, no changes whatsoever, thats why only "Super fast" is included. But what did remove the ufo trailing dramaticaly is lower contrast. Reducing it removes trailing completely but the overall motion blur still felt the same with either improved or default osd settings.
The thing is, with this specific monitor contrast acts the same as brightness so you can change it as much as you want and you wouldn't be able to tell whether the brightness or contrast is changed.

NVIDIA GPU: (ufos from The Solid Lad, unknown nvidia 10se gpu)
Default OD mode for the AW2518HF is "normal" and it looks the same as improved contrast settings on amd gpu. When u change from "Normal" to "Fast" or "Super fast" it introduces corona/white trailing behind ufos. I don't know if reducing contrast fixes this artifact but I believe it does not.

NON-SUPPORTED ADAPTIVE SYNC GPU: (tested myself with R7 370)
The corona on "Normal" mode is slightly visible but noticeable. "Fast" is very present (same as nvidia gpu "Fast") and "Super Fast" is extremely distracting and unusable (worse than on nvidia gpu "Super fast"). But this time over drive modes felt slightly less blurry in games wich means response times are reduced unlike with rx 580 od modes. Reducing contrast also did not change the corona artifact.
I do not know how monitor behaves on non-supported AS nvidia gpus, probably the same.


AW2518HF @240Hz | ufo speed 960px/s (24,5" 1080p)
all_2_jpg.jpg
all_2_jpg.jpg (418.74 KiB) Viewed 20293 times
non-s_3_jpg.jpg
non-s_3_jpg.jpg (95.62 KiB) Viewed 20293 times
Last edited by Discorz on 14 Apr 2020, 09:33, edited 1 time in total.
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

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

Re: ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Post by Chief Blur Buster » 09 Apr 2020, 12:59

Fascinating pursuit camera work -- I assume you did the pursuit camera -- very good job!

(Would it be possible to link to the originals? I love seeing the original sync track -- the "horizontal ladder" you see at www.testufo.com/ghosting -- because the photographed sync track (to me) is literally almost as golden as a cryptographic certificate of pursuit camera accuracy)
Discorz wrote:
08 Apr 2020, 10:27
But what did remove the ufo trailing dramaticaly is lower contrast. Reducing it removes trailing completely but the overall motion blur still felt the same with either improved or default osd settings.
This happens with most overdrive -- good overshoot room below black and overshoot room above white -- helps reduces ghosting/coronas.

It's also a known strobe-crosstalk-reduction fix too on some models of gaming monitors -- the contrast-ratio reduction trick. Old LightBoost monitors did this partially to reduce 3D glasses crosstalk. This technique still works today when done manually via reducing contrast while simultaneously using slightly stronger overdrive (since the reduced contrast makes stronger overdrive work better and more cleanly).

Fortunately, GtG is fast enough -- and VBIs large enough -- that we don't necessarily need to do this trick anymore -- giving up contrast ratio (LightBoost-style crappy colors) in order to get low-crosstalk strobing. For example, the ViewSonic XG270 (240Hz 1ms IPS panel), for example, can now strobe relatively cleanly at 120Hz without requiring a contrast-ratio reduction.
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!

User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Post by Discorz » 10 Apr 2020, 08:40

Thanks. My test were only the top 3 and last one. Nvidia's ones I found on yt.

Here are the originals (22MB "png"):
https://mega.nz/file/o4RkyShB#8Bf5g_SUw ... 6Jtnu-OdQ8

I thought people should know this. They might see online that the monitor has this artifact and then avoid buying it. But in fact monitor is extremely good.

Another interesting thing about this monitor; Reducing the gain for green color (customizing colors in osd) from default 100 to 99 also removes the ghosting significantly. All other settings are default. Same thing with contrast even from default value of 75 to 74 does the trick. By going lower ghosting keeps improving.
Theres so much to learn about this, every day I keep seeing new things.
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

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

Re: ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Post by Chief Blur Buster » 10 Apr 2020, 22:08

Neat to cross-compare your pursuit camera test to a somebody elses' test results!

(In this case, A Solid Lad's YouTube freezeframes. That YouTuber is a forum member here, too)

Good image-colors equalization & alignment job for easier visual comparison of WYSIWYG ghosting artifacts.

So great job, I've seen worse error margins in same-reviewer two-display comparisions before...

<Advanced pursuit camera commentary>
Giving a allowance for error margins (pursuit camera tracking error chart / pursuit camera sync track artifacts allowance) -- this is a relatively accurate cross-comparision between two different people doing pursuit camera with different equipment.

Unlike comparing input lag tests across different websites (which is much more unreliable yet due to complexity of lag tests and different stopwatching criteria varying from GtG2%, GtG50% and GtG100% on different websites, and different lags of different colors and different sync technologies, and the complexity of latency gradients where not all pixels have identical input lag) -- it is feasible and mostly reliable to compare the UFO pursuit camera tests across different testers when within sufficient error margins and assuming colors/histogram are adequately equalized.

That said, more subtle differences like GtG 0.5ms vs GtG 1.0ms will require much more accurate tracking that may require a camera rail to eliminate the camera-tracking-blur error margin -- though sometimes video practice, I've seen some people achieve "chance" 0.25ms hand-wave accuracies (1 in 1,000 to 1 in 10,000 video freezeframes). Fortunately, the sheer freezeframe spray of video pursuits allows such overkill of scientific samples -- sometimes gives enough hand-wave samples that single chance samples match rail accuracy. After all, only 1 frame needs to be accurate, no matter how many attempts you do.

This is one of these times where I'd happily fund a Blur Busters X-Prize for a script / program / etc that automatically scans a mostly-shaky video file to find the clearest pursuit-camera freezeframe, via AI recognition of the sync track and finding the most-aligned tickmarks via AI algorithm --

There is aspirations to create a pursuit camera app that works on smartphones -- you keep hand-waving smartphone until it beeps, and you've got a shareable pursuit camera freezeframe (automagically chosen by the app by AI or other algorithm)
</Advanced pursuit camera commentary>

___

Back on topic -- mind if I ask what camera you used, and did you use photo technique or freezeframe-from-video technique?

I presume you used the easy rail-less hand-wave pursuit camera for your top 3 and the last one?
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!

User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Post by Discorz » 14 Apr 2020, 11:21

Its just a basic smartphone camera (ZTE Blade A610), hand recorded and freeze frame technique used. So all 3 are same. It kinda worked pretty well. There was no color eqing needed. Just had to scale it properly. "Home made" ufos would probably be just fine for most people. Something is better than nothing.

I guess from now on I'll include the sync tracks if I ever buy a new monitor. I'm very curious how those new IPS 240-280Hz panels perform and especially that new ASUS's 360Hz TN montor. I feel like they can push it much further with refreshrate but they won't give it to us for some reason. I'd buy 480Hz even if its a bad color reproduction.
Also, when will very fast (≈3000-3840 px/s) ufos be a standard? That would be a more realistic speed for gaming. 960 px/s is more for average desktop usage I guess. Because even on good 240Hz monitor I noticed I still had to stop moving mouse around to avoid motion blur and see clearly (even tho ufos on 960px/s show it clear). It would be nice to have some ufos representing ingame motion blur.

Blur Buster app for pursuit would be awesome. If u ever make this happen let people share their ufos and collect it all in one searchable database. So if people need to check the response times of a specific monitor they can just go there, type the monitor name and check it out. Something like this would be awesome. You guys would probably just have to approve all those home made pursuits :D.
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

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

Re: ALIENWARE AW2518HF UFOs | AMD (freesync) vs NVIDIA (g-sync compatible) GPUs

Post by Chief Blur Buster » 14 Apr 2020, 13:48

No color EQ? Wow.

Smartphones are getting better and better -- even the cheap ones (at least those made recently)

The main weak link would be AI algorithms necessary to recognize the sync track (my invention) and automatically recognize the common sync track alignment issues. Technology is rapidly catching up to the point where this is beginning to be practical, with the right skills.
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!

Post Reply