Blur Busters Forums

Chief Blur Buster wrote: ↑

24 Aug 2020, 12:08

KingMusti wrote: ↑

23 Aug 2020, 13:15

Hi, I wanted to ask if the results are good, once with G-Sync and once with LG MBR. my monitor is the lg 27gk750b-f

Hey,

Great photos! Did you use stationary camera (OK for strobe crosstalk, if exposure is limited to exactly one refresh cycle) or pursuit camera (best for universal display motion blur photography).

For more accurate strobe crosstalk photography, please use www.testufo.com/ghosting and follow the handwave smartphone instructions. The sync track makes it easier to tell whether you pursuited correctly or not.

[Moderator Note -- Fixed IMGUR URL]

May I ask if there's a point overclock the vertical total if I'm using freesync? Does it make a difference?

daniloberserk wrote: ↑

20 Dec 2020, 14:47

May I ask if there's a point overclock the vertical total if I'm using freesync? Does it make a difference?

AFAIK, you'll get marginally lower display lag near the center of your screen.

A Solid lad wrote: ↑

22 Dec 2020, 06:44

daniloberserk wrote: ↑

20 Dec 2020, 14:47

May I ask if there's a point overclock the vertical total if I'm using freesync? Does it make a difference?

AFAIK, you'll get marginally lower display lag near the center of your screen.

Thanks! For some reason when I OC my display with CRU it doesn't detect as an freesync display. So I'll probably not ocing for now =)

Overclocked vertical total is useful for strobe crosstalk reductions and/or latency reductions.

That said, decision whether to overclock the vertical total OR overclock the refresh rate, depends on panel capabilities. For example, max-unoverclocked-Hz with overclocked Vertical Total, sometimes means the monitor can handle an overclocked Hz with a reduced vertical total.

Monitors does not always work this way especially if they use TCONs that scan-converts (see this Quick Frame Transport related post where some panels refresh at a different scanout velocity than the cable), and monitors can sometimes pop up an unwanted symbollic "OUT OF RANGE" during Hz overclocking even if the panel theoretically works overclocked just fine.

Chief Blur Buster wrote: ↑

22 Dec 2020, 18:15

Overclocked vertical total is useful for strobe crosstalk reductions and/or latency reductions.

That said, decision whether to overclock the vertical total OR overclock the refresh rate, depends on panel capabilities. For example, max-unoverclocked-Hz with overclocked Vertical Total, sometimes means the monitor can handle an overclocked Hz with a reduced vertical total.

Monitors does not always work this way especially if they use TCONs that scan-converts (see this Quick Frame Transport related post where some panels refresh at a different scanout velocity than the cable), and monitors can sometimes pop up an unwanted symbollic "OUT OF RANGE" during Hz overclocking even if the panel theoretically works overclocked just fine.

Thank you very much Chief! I'll try to OC again later, just need to find why it just doesn't detect Freesync on the AMD Driver. I just love all the knowledge and kindness that people on this community have to help! Happy Holidays!

so I followed the advice in this thread to set up a custom res with CRU to eliminate the stutter issue with ULMB. The issue is that I play games at 1728x1080 stretched (16:10, used to play 1680x1050 but made a custom res for this) and with stretched res I'm getting the stutter issue. Can anyone give advice on how to set up my custom res to eliminate the stutter? Obviously using the same pixel timings for 1920x1080 doesn't work for the smaller res.

Some ULMB monitors have incompatibility with monitor scaling with ULMB. Use GPU scaling instead with ULMB.

Chief Blur Buster wrote: ↑

13 Jan 2021, 22:35

Some ULMB monitors have incompatibility with monitor scaling with ULMB. Use GPU scaling instead with ULMB.

Hey Chief! Recently I've watched a FR33THY's video about display scaling. He advocates that we should use display scaling if we have a monitor compatible with it, claiming that it brings a lot less of input lag. You still think that GPU scaling should be used? I have a GTX 1080TI and don't intend to use ULMB, just 240 with uncapped FPS in CS:GO (only game I play).
If I want to create a resolution (i.e. 1280x960), should I worry with another setting in CRU or just make sure that I can get to 239.760hz? I've created a few resolutions by changing the active resolution that I want and then changing "Total" horizontal pixels and verical lines till I get exactly 239.760. That is the correct way to do it?

dev1lwearsprada wrote: ↑

14 Feb 2021, 16:28

Hey Chief! Recently I've watched a FR33THY's video about display scaling. He advocates that we should use display scaling if we have a monitor compatible with it, claiming that it brings a lot less of input lag. You still think that GPU scaling should be used? I have a GTX 1080TI and don't intend to use ULMB, just 240 with uncapped FPS in CS:GO (only game I play).

Depends.

It's not universal advice. This was usually mostly true in the old days, but with faster GPUs, there are cases where GPU scaling outperforms display scaling. So with one computer, you have less lag via display scaling, but with another computer+display, you've got less lag via GPU scaling.

Pseudocode:

IF (GPU = fast & recent) AND (Display = has slightly laggy scaler) THEN use GPU scaling.

However, you'd probably need an LDAT or other similar tool to confirm which is better for your setup.

In your case, you've got an incompatibility consideration (ULMB mode incompatible with monitor scaling) and so you have no choice but to use GPU scaling on some models of displays when using G-SYNC or ULMB.

So that's the two answers.
(A) The "depends" factor where you might have less/more lag via GPU scaling or display scaling.
(B) The "compatibility" factor that prevents monitor scaling (e.g. a specific mode).