XL2720z V2 + BB Strobe Utility = Greatest Gaming Monitor!

[Chief Blur Buster]

I've found that you can avoid the pixel clock patch for the VT1350 tweak by using a Horizontal Total 2040 instead of 2080. That bumps the dotclock just barely below the limit.

[Magicson]

Thanks guys! Should i give a try to 2040/1350@120hz or patch and go for 2080/1350@120hz?
I will test after work, will let u know how it went!

[Magicson]

I've tested and it didnt help at all. I'm still getting LOTS of crosstalk in the WHOLE screen. Maybe i'm doing something wrong

[Falkentyne]

Crosstalk doesn't cover the entire screen, even without tweaks.
Are you talking about GHOSTING?

Crosstalk is a very thick inverse band that is far more pronounced than ghosting.
Ghosting is simply trailing artifacts behind the image, without AMA (AMA turned off), they are double or triple image copies (primary and secondary trail, with the primary trail being brighter and the secondary being faint).
With AMA (RTA/overdrive) turned on high (change this BEFORE enabling blur reduction as there are preset values that get overwritten for blur reduction's own calibrated values), you get an inverse to a transparent primary trail, and a faint secondary trail.

Crosstalk is completely different.
Crosstalk is a VERY thick inverse band that actually SHIFTS the main image forward or behind a frame, due to the GtG transitions either settling before the beginning or the end of the strobe pulse (this is very hard to explain, but theres a better explanation here: http://display-corner.epfl.ch/index.php/BenQ_XL2411Z ). To see the difference, enable blur reduction at 100hz (just in case you're using internet explorer), restart IE just incase, go to testufo, moving photo, alien invasion, you should see a perfectly smooth 100 fps image. Now use the blur busters' utility and move the strobe phase slider slowly from left to right.

You will see the crosstalk band move from the top to the bottom and you will notice the "main" image actually SHIFT a frame to the right by a frame (strobe phase set to high) or to the left by a frame (strobe phase set to 0).

So as you can see, it's completely impossible for the crosstalk band to cover the entire screen. At MOST, it can cover one third of the screen.
With a VT 1500 tweak and the strobe phase set to "0", at a persistence of 1.0ms or lower (strobe duty 006), the crosstalk band should be at the very top first UFO field only (set to fullscreen on the test, with the F11 fullscreen also on the browser to remove all of the menus), with 'reverse' overdrive trailing right below that, then becoming almost transparent trailing in the middle, then at the bottom, where the settling errors are, double/triple primary/secondary copies (not inverse or as transparent) around the bottom (basically, the 'bottom' crosstalk band which pushes the image ahead a frame, is off screen if you are using the VT tweaks). If you were NOT using the VT tweaks, at strobe phase 000, the top of the image would be extremely clear while the middle to the bottom would have the crosstalk field there.

[Magicson]

Crosstalk doesn't cover the entire screen, even without tweaks.
Are you talking about GHOSTING?

Crosstalk is a very thick inverse band that is far more pronounced than ghosting.
Ghosting is simply trailing artifacts behind the image, without AMA (AMA turned off), they are double or triple image copies (primary and secondary trail, with the primary trail being brighter and the secondary being faint).
With AMA (RTA/overdrive) turned on high (change this BEFORE enabling blur reduction as there are preset values that get overwritten for blur reduction's own calibrated values), you get an inverse to a transparent primary trail, and a faint secondary trail.

Crosstalk is completely different.
Crosstalk is a VERY thick inverse band that actually SHIFTS the main image forward or behind a frame, due to the GtG transitions either settling before the beginning or the end of the strobe pulse (this is very hard to explain, but theres a better explanation here: http://display-corner.epfl.ch/index.php/BenQ_XL2411Z ). To see the difference, enable blur reduction at 100hz (just in case you're using internet explorer), restart IE just incase, go to testufo, moving photo, alien invasion, you should see a perfectly smooth 100 fps image. Now use the blur busters' utility and move the strobe phase slider slowly from left to right.

You will see the crosstalk band move from the top to the bottom and you will notice the "main" image actually SHIFT a frame to the right by a frame (strobe phase set to high) or to the left by a frame (strobe phase set to 0).

So as you can see, it's completely impossible for the crosstalk band to cover the entire screen. At MOST, it can cover one third of the screen.
With a VT 1500 tweak and the strobe phase set to "0", at a persistence of 1.0ms or lower (strobe duty 006), the crosstalk band should be at the very top first UFO field only (set to fullscreen on the test, with the F11 fullscreen also on the browser to remove all of the menus), with 'reverse' overdrive trailing right below that, then becoming almost transparent trailing in the middle, then at the bottom, where the settling errors are, double/triple primary/secondary copies (not inverse or as transparent) around the bottom (basically, the 'bottom' crosstalk band which pushes the image ahead a frame, is off screen if you are using the VT tweaks). If you were NOT using the VT tweaks, at strobe phase 000, the top of the image would be extremely clear while the middle to the bottom would have the crosstalk field there.

Thanks! I'll test that later. Asuming it is ghosting and not crosstalk, how can i reduce it? I've googled it, and it seems that lightboost seems to have less ghosting, can i use lightboost on my XL2720Z? is it worth it?
Thank you so much for your time/knowledge

[Falkentyne]

You can't reduce it much, although if you're using contrast 50, try lowering it to 40-43; that will improve the ghosting at the lower part of the screen slightly.

You can set the contrast to 0, which will improve the bottom of the screen significantly (close to lightboost quality) while it won't affect the top (the inverse ghosting doesn't get affected; the transparent ghosting becomes almost invisible).

What's interesting is that lightboost's accelerated scanout seems to have the strobe pulse on (under) the bottom instead of the top.
In lightboost mode (on both the Benq and the asus VG248QE), you will notice the inverse ghosting at the -bottom-, with the top being crystal clear, while in benq blur reduction mode (with contrast set to 0 and with all VT tweaks), you notice the inverse ghosting at the top with it being close to clear at the bottom.
It's because of this, why Benq used 100 strobe phase as default, (which brings the 'transparent' part, which can be improved by contrast lower, to the top), to match lightboost's setting. But since BBR doesn't have accelerated scanout (which would make the BOTTOM of the screen like lightboost (inverse ghosting instead of big crosstalk field), you get the crosstalk mess. However you can not use strobe phase 100 with the VT tweak, because once you start moving the crosstalk field down (trying to make it like lightboost), after strobe phase 49-54 (depending on ur refresh rate), the strobe phase stops changing and the persistence starts going down by itself (although the strobe duty value doesn't change), until you reach strobe phase 60, where the backlight shuts off (don't worry, this can only be done through the service menu; the windows utility won't exceed 47)

All of the 27" screens have more ghosting than the 24" screens. This goes for both lightboost and BBR.
The tradeoff is that the 27" screens have better contrast in return, in lightboost mode, and the colors don't take such a hit as they do on the 24" screens. the 27" benq's can maintain almost 1000:1 contrast in lightboost mode at 50 contrast (you will have the purple lightboost tint, though, if you go past 37 contrast), which is MILES better than the VG248QE. Almost every other lightboost monitor (all the 24" ones) can BARELY reach 500:1 in lightboost mode. The ghosting will be much worse than the VG248 unless you lower the contrast to 0, which seems to match the contrast ratio of the VG248 even when the VG248 is set to around 40 or so. But at contrast 0, while the Benq then looks almost as good as the VG248QE's lightboost, (almost no ghosting in at all except the settling errors at the bottom) but ...well...you get the washed out image that the VG248QE got bashed on in lightboost mode for...

Yes, the ghosting IS less in lightboost mode compared to BBR mode, because lightboost has per line overdrive due to the LC panel, and the crosstalk is nonexistent (its basically identical to VT 1500, with strobe phase 0 and strobe duty (persistence) at 002 (0.333ms), pretty much identical to that, because lightboost uses accelerated scanout, and the VT tweak at 1500 with duty 0 and persistence 0.333 matches up with that (you can only go lower than 0.5ms persistence through the service menu).

You can lower the ghosting slightly in Benq BR mode (the article I linked from strobemaster in the above post also mentions this!) by changing contrast 50 to 40. I personally use 43 contrast as that matches the grayscale point spot on. If you drop it all the way down to 0, it becomes better but you lose a lot of the strengths of the 27".

The only way to get lightboost mode on the Z series is by using an Nvidia card with the INF override. I used the INF override combined with the 3d vision emulator, with using a DVI to HDMI adapter and focusing a custom 1149VT 1920x1080 120hz refresh rate through my laptop's HDMI port, and unlocked lightboost that way. I believe the 3d vision emulator.zip file (3d vision discover) worked with the INF override the easist, but apparently the emulator doesn't work on newer NVidia drivers after 310.xx. If you have an AMD card, you need an Nvidia card/laptop as a surrogate to unlock lightboost.

[Chief Blur Buster]

Crosstalk and ghosting are actually interlinked!

They cannot be 100% completely separate; though they are different terminologies to describe different artifacts, that are caused by the same root cause (aka LCD response!)

Crosstalk is the bleeding of refresh between strobes creating a double image effect even during framerate-refreshrate matched motion (e.g. 120fps@120Hz creating double images).

Crosstalk artifacts can also includes bleeding of ghosting/artifacts not successfully hidden in the total darkness between strobes. Normally, a strobe backlight & LCD overdrive, is calibrated in such a way, so that as much of the ugly stuff (ghosting and overdrive artifacts) are hidden from human visibility in the black periods -- the off cycles -- between flashes of a motion blur reducing strobe backlight.

For more information, see:
LCD Motion Artifacts 101
LCD Overdrive Artifacts
High Speed Video Of A Strobe Backlight (also applies to BENQ BR)

Also, see Motion Blur Photography Examples which contains good photographs how strobe backlights actually "chop-up" an LCD ghosting artifact (crosstalk between refresh cycles), as one side effect of combining a strobe backlight with an LCD that's not fully calibrated to hide artifacts in total darkness between refresh cycles:

Ghosting without strobing

- Lots of motion blur, and you see a ghost trail to the left.

Ghosting as part of Strobe Crosstalk

- No motion blur, but the existing ghost trail is now "chopped up" in the form of strobe crosstalk.

As you can see, motion blur reduction can come with the side effect of the existing ghosting being "chopped" (which I call strobe crosstalk -- LCD ghosting is also a cause of 3D crosstalk too: Unfinished LCD pixel response that still shows up at the next strobe flash). What we call "3D crosstalk" during stereoscopic 3D will also be visible during 2D motion as strobe crosstalk (sharp ghosting during fast 2D motion).

The faster an LCD is (and the better the overdrive), the more the LCD can hide the ghosting before the next strobe flash.

For vision and optical illusion geeks: Ever noticed the dotted trail effect on flickering light sources? Same thing. For non-display cases, the dotted trail effect of a flickering light in motion (e.g. PWM-driven LED car tailights, while driving at night, and flitting your eyes around -- or red LED alarm clocks at night -- or an orange neon wall nightlight) -- the dotted trail is also known as the "phantom array" effect, a stroboscopic effect, where a fast-moving flashing light creates a dotted trail, seen at the bottom of Scientific References. Different retina position during subsequent flashes of light, creates the multiple-image trail.

[qqbr]

Hello Chief. I need a little help. I changed my videocard to R9-280X how to enabled tweak 1502 and 128hz
before i use GTX680 and no problem w this tweak.
Thanks

[Falkentyne]

128hz may be problematic with 1502 VT.
You're approaching the absolute limits of what you can do over DVI.
I suspect 128hz will have the same problem as 76 hz...(76hz is the cutoff between double strobe being allowed for strobing and single strobe being always forced on. I think that this cutoff is why 76hz goes out of range half the time with a VT 1500 tweak and the other half of the time it displays an image.

Why not just do 120hz?
with lower timings of 2080 HT and 1500 VT, and front porch 48,3 and sync width 32,5 you're already at 374 MHz bandwidth.

To enable VT tweaks just use custom resolution utility.
This works perfectly: ^^ the timings above for 120hz.
you will need the toastyX pixel clock patcher since you are exceeding 330.00 MHz bandwidth.

You can also do 2200 HT, 1500 VT, 100hz, front porch 88/4, sync width 44/5, for 330.00 MHz exact bandwidth at 100hz refresh rate, won't need patcher for that.

You can restart the driver without rebooting by using the driver restarter, after any changes made in Custom resolution utility!

If you're using the omega drivers and the amd CCC does not automatically restart after running "restart64.exe" (toastyx driver restarter also on the monitortests forum along with CRU), just right click the desktop and load CCC.

[qqbr]

I have for a long time use 128hz and VT 1502 it works fine w me. on GTX 680 absolutely no problem