Page 2 of 2

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 11:24
by Vega
Chief Blur Buster wrote:
Vega wrote:Well this test was a bust. Cross-talk is WAY worse on the Samsung's than ULMB. Funny thing is that the 4-zone scan makes the cross-talk uniform vertical across the panel. The problem is it's bad cross-talk. :x
P.S. It's strobe crosstalk. I add strobe as a qualifier because the word "crosstalk" comes from the use of 3D Vision glasses.

I am surprised because a tester said one of the Samsungs did better than ULMB.
Perhaps it is something that varies a bit... So I'm curious if there's a variance somewhere.

Did you warm up the LCD for at least 30 minutes?

Cold LCD's ghost much more than warm. Especially VA panels.

EIZO FG2421 ghosting, when cold:

Image

EIZO FG2421 ghosting, after warming up for 30 minutes:

Image

So warming up (higher temperatures) makes a really huge difference on VA panels.
Ya, I had the monitor on for a good hour. Was really surprised at how bad the strobe crosstalk was. Pretty much the worst strobing backlight I've seen. I double and triple checked settings. I could definitely tell it was strobing. The strobe crosstalk was pretty uniform (which is a benefit over ULMB which is usually just minimized in the center), but overall was quite a bit worse than my SG2716G in ULMB clarity.

Who knows, maybe the 27" Samsung version I have is much worse than the 24" for whatever reason?

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 11:42
by Chief Blur Buster
Vega wrote:Ya, I had the monitor on for a good hour. Was really surprised at how bad the strobe crosstalk was. Pretty much the worst strobing backlight I've seen. I double and triple checked settings. I could definitely tell it was strobing. The strobe crosstalk was pretty uniform (which is a benefit over ULMB which is usually just minimized in the center), but overall was quite a bit worse than my SG2716G in ULMB clarity.

Who knows, maybe the 27" Samsung version I have is much worse than the 24" for whatever reason?
Strange -- someone will need to do official tests.

I had heard that these monitors did an excellent job.

Maybe it's a firmware thing? You said it was a test monitor -- did Samsung send it directly to you?

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 12:54
by Vega
It was a January build, so I would think the firmware wouldn't have been too old on it. For test, I just meant I ordered one to test to see if they future 2017's would be viable for me. I had no plan on sticking with this sample monitor since it's only 1080p.

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 17:45
by Chief Blur Buster
Darnit!

I guess the verdict now becomes...."inconclusive".
I hear from 2 other people saying that there's near-zero strobe crosstalk on the Sony monitors.

BTW, we now have the budget to finally purchase several new "Better Than 60Hz" monitors this year -- so we'll finally be doing real tests (of even higher quality than our G-SYNC Preview Articles #1/#2). Keep tuned!

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 18:07
by Falkentyne
Vega wrote: Who knows, maybe the 27" Samsung version I have is much worse than the 24" for whatever reason?
This wouldn't surprise me. The 24" Samsung was reported as having basically no crosstalk at all.

Remember Lightboost mode on the 24" monitors and the 27" 1080p monitors?
Lightboost on the 24" monitors using the AUO M240HW01 panel looked excellent (the VG248QE looked the best, followed by both the T And Z series Benqs (if Lightboost were unlocked). The 27" monitors had a lot worse ghosting than the 24" monitors, and you would usually have to drop the contrast down to 0 for it to look decent, but it still showed more ghosting.

Could be the same thing here just regarding the panel size difference.

I'm too poor to run around buying and testing monitors...so I can't be of much use. I still haven't found a monitor to replace my XL2720Z. Still trying to consider the XL2735 if it drops down in price, so I can see if freesync can work on it via ToastyX CRU.

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 20:50
by Vega
Oh that is one thing I didn't test was messing with the contrast on "black equalizer' settings during strobing. I guess I can see if those adjust the strobe cross-talk much.

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 29 Mar 2017, 21:16
by Vega
Samsung on fastest 1ms MPRT 144 Hz strobe:

Image


S2716DG 120 Hz ULMB "100 brightness".

Image

You can see the Dell destroys the Samsung. Identical camera settings.


Could it simply be TN>VA for strobe backlight?

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 30 Mar 2017, 01:28
by Chief Blur Buster
I've seen far worse 144Hz strobing. That's why NVIDIA doesn't disable ULMB for 144Hz.
Strobing doesn't work very well with reduced vertical totals. ULMB automatically internally uses (automatically, without user intervention) the equivalent of Large Vertical Totals during 120Hz -- to create a larger blanking interval to have a longer pause for LCD GtG transitions in darkness before strobe backlight flash.

Have you tried Blur Reduction at 120Hz? That probably looks better on the Samsung.

From German site Lim's Cave:
Image

Not sure if that's 120 or 144, but it seems to look better...

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 30 Mar 2017, 05:32
by Vega
Ya 120 Hz didn't look any better. Notice the spaceship white "belt" dots on my samples space-ship smear all the way to the next vertical divider. It's a little more evident in real life viewing. That is a lot of extra strobe-crosstalk that removes a lot of the motion clarity benefit IMO.

I wonder why that German site appears to have a clearer strobing Samsung than me. Also, contrast and "black equalizer" settings have hardly any affect on the strobe cross-talk trailing. As most know, the "brightness" setting is locked out while strobing on these Samsung. Really stupid decision.

Re: Samsung's 2017 Scanning Back-Light Monitor Lineup

Posted: 30 Mar 2017, 11:30
by Chief Blur Buster
Interesting -- I think we'll need more data whether this is a panel lottery, model difference, firmware, monitor OSD configuration issue, etc...