My ViewSonic XG270 + PureXP™ Tweaks (ICC Download) [Advanced Users]

[Chief Blur Buster]

Your numbers are different. Double check the numbers.

Also:
Vertical Total must be bigger than sum of three numbers above it
(NVIDIA Control Panel autocalculates Back Porch based on the rest of the numbers)

Front Porch must be smaller than Total
Back Porch must be smaller than Total
Sync must be smaller than Total.

Video signals are structured like this:



Total = Active Resolution + Back Porch + Sync + Front Porch

Copy the numbers again, because not all numbers match the other screenshots.

[donnie]

Not sure what am I doing wrong:



NVCP screen on the right is exact copy of instructions from your 120hz low lag trick. It works. But trying the same in CRU doesn't work as it doesn't accept the front porch 1080 trick. Do I need to convert pixels\lines or what?

[ELK]

Not sure what am I doing wrong:



NVCP screen on the right is exact copy of instructions from your 120hz low lag trick. It works. But trying the same in CRU doesn't work as it doesn't accept the front porch 1080 trick. Do I need to convert pixels\lines or what?

I'm not 100% sure if you must delete the CEA-861 block. something about nvidia and default extension blocks maybe... Anyways I deleted the CEA-861 and added a new DisplayID block. I then added a detailed resolution to the DisplayID.

DisplayID will accept large front porches.

I originally learned this so that I could make my headless monitor run 4k@90hz (There is no monitor it's just a display port connection, with no cable, and and monitor information build it so that you can use any resolution in Virtual desktop, or bigscreen, both VR applications) but I use it now for these QFT resolutions on my actual monitor. The XG270.

New Bitmap Image (16).png (138.59 KiB) Viewed 8677 times

This one shows you that 120hz is possible. The exact timing you are aiming for.

New Bitmap Image (17).png (75.46 KiB) Viewed 8672 times

[donnie]

That seems to work thanks, what happens if I add DisplayID block additionally to CEA-861?

[Chief Blur Buster]

Good point about NVIDIA vs CRU inflexibility.

About last time...
Technically, front porch vs back porch did not seem to have an effect on latency tests with QFT.

Originally I thought this was something I could use to make RTSS Scanline Sync unnecessary for QFT.
I thought Windows frame presentation API Present() blockage delayed API return to the VSYNC portion of the VSYNC signal, but it apparently pageflips at the beginning of the whole shebang of a VBI (right after last pixel row). So I still needed RTSS Scanline Sync to force end-of-VBI pageflip to get the the inputdelay needed to gain the Quick Frame Transport (QFT) benefits of VSYNC ON / reduced strobe lag.

Now, I need to do more testing to see if latency results improve with various NULL settings (NVIDIA Ultra Low Latency) and see if the NVIDIA drivers will change Present() blockage. This is part of the magic sauce of "Quick Frame Transport" but most less-experienced employees at Microsoft or NVIDIA doesn't understand display refresh scanout well enough to realize input latency reductions during QFT requires this change.

We may begin to publish Blur Busters articles about Quick Frame Transport on the main Blur Busters website in the near future.

A much easier Quick Frame Transport would theoretically be achieved this way, with only a minor setting change to a graphics driver, while simultaneously enabling VSYNC ON + NVIDIA's Ultra Low Latency feature. This becomes a much easier substitute than RTSS Scanline Sync for the average user.

Many new high-Hz monitors can accept lower-Hz frame transmission faster than its refresh cycle -- like a 120Hz refresh cycle transmitted in 1/240sec -- and this allows a Large Vertical Total to hit two birds with one stone (simultaneously reduce strobe crosstalk AND reduce strobe lag).

For now, the mandatory requirement of RTSS Scanline Sync is needed, because QFT increases lag with many graphics driver settings (beginning-of-VBI frame presentation increases lag, end-of-VBI frame presentation decreases VSYNC ON lag)

[ELK]

That seems to work thanks, what happens if I add DisplayID block additionally to CEA-861?

Then the DisplayID resolutions will not show up. I haven't tested this in awhile though. try adding DisplayID in addition to the CEA-861 if the resolutions show up then everything is good

[Kuzunkhaa]

Thanks for the ICC tuned for strobed mode! Could you add the contrast your xg270 is set to? It's in Display -> Imagine adjust -> contrast.

Hello ELK,
I'm so sorry for the late response, but these days I was really busy and I didn't check the forum.

Anyway the color calibration was made with the predefinite value 70, because in LCD display and also suggested in DisplayCal software, is better to maintain the factory default level to not mess up "the black point" and "the color temperature".
Sure, probably we can do a little bit of tuning for that value, but I've reached the sRGB 99.7% in this way and I think this was fine for us
Enjoy your monitor and have fun!

[Otzl]

Hi, how would I go about making the DisplayID resolution active? It's not possible to do the 1080 front porch trick without creating a detailed resolution under DisplayID, but I'm not sure what to do next. How do I use it? Also, is the colour supposed to be 6 or 8 bit after doing these tricks?

Thanks in advance.

[Gregix]

Great guide man, handy, but AWARE those tryin to use this ICC profile. That can, but may not work.
As matter of fact, I just bought this same monitor, updated firmware, and at start to get lowest dE at begining of measurment I had to set my RGB values 83, 85, 87 and set contrast to 81. And even with that from start my dE was 0.5, not 0.1 like OT.
I just had from start not enough blue, and it tinted my picture little too warm, was obvious to me btw on first look(and as compared to my sec display, calibrated Benq xl2546).
So
Same x-rite device, same software used, same screen, different results. AAfter calibration my my luminance is lower(142.38cd/m2), WP is 6530K, 99,7sRGB coverage now.