BenQ XL2720T: 144hz 1920x1080 Matte Nvidia 3D Vision 2 Gaming Monitor

Zero-7 said:
XL2720T lightboost hack, you will not notice in game, because is super smooth, no blur.
Click to expand...
On all LightBoost monitors, including the BENQ XL2411T (which is a slightly faster responding LCD than the XL2720T), I always clearly see a difference when LightBoost is enabled versus off. Especially when I enable VSYNC ON for solo gaming, non-competitive gameplay. (which yields the same 'perfect smooth motion' that PixPerAn has, LightBoost is easier to notice when you enable VSYNC ON)

Especially:
- Strafing left/right in front of highly detailed textures
- Circle strafing
- Fast 180 degree turns
- Running while looking at sides, floor, ceiling, etc.
- Seeing stuff zoom by (racing, flybys, etc)
- Etc.

For my eyes, it's quite pretty clear, and I can see the difference between LightBoost=10% versus LightBoost=100%, especially when I enable VSYNC and I'm running at a perfectly synced 120fps@120Hz -- easy to do in Source Engine games. Standing in front of a high-resolution texture wall (e.g. standing 5 virtual feet in front of a high-resolution wall with a poster), and then strafing left/right. In situations like this, the difference between LightBoost=10%, LightBoost=50%, and LightBoost=100% becomes noticeable. My eyes is very sensitive to motion blur (part of the reason I enable VSYNC during solo gameplay, because stutters hurt the LightBoost effect, when the input lag is not as important -- though I turn VSYNC OFF when playing competitively).

If you disable VSYNC, then the difference between LightBoost=10% versus LightBoost=100% is almost diminished by non-synchronization of fps vs Hz, although it becomes noticeable again in source engine games if you really uncap your framerate (e.g. fps_max 999) to eliminate unexpected harmonics between fps and Hz. (fps_max 250 = harmonic of 10 Hz = 10 micro stutters per second). Even a GTX 680 can run at 250fps inside closed spaces in a source engine game, so it's very easy to get unexpected stutters caused by harmonics between fps and Hz. For motion-blur-sensitive people, stutters produce a very tiny amount of motion blur, which can cancel out the difference between LightBoost=10% versus LightBoost=100%.

Generally, I play at LightBoost=50% now, since LightBoost=10% is too dim during daytime.

Everybody's eyes is different.
 
Last edited:
David.Garcia said:
Out of those Asus monitors which is the latest? why is it that only the asus 24" has displayport whereas the 27" dont OMGT_T
Click to expand...

VG278HR is the latest.

Yea. Just saw the specs on that for the 27s..no display port...sad that is
 
Last edited:
But, what some of us need is these questions answered:

What is the best Light Boost monitor out there right now?

Is there a better 27 inch over other 27 inchers---Benq vs Asus?

Is 24 inch Asus better than the 24 inch Benq?


Does anyone have a price for the ASUS VG278HR?
 
I'm itching to buy this monitor but was waiting on more reviews to pop up. I'm a little worried by the disclaimer on PC Monitors. I don't want a run of the monitor that has 'issues'.
 
All the 27"ers use the same panel that will double ghost in Lightboost.
 
You will see the double ghost effect in desktop mode. You wont be able to see it in gaming. (Well I never did, and I tried very hard to look for it.) I never notice the effect until Vega brought it up to my attention. I think if we can disable lightboost in desktop and go into lightboost automatically in gaming, that would be a good compromise.
 
Mark Rejhon said:
If you disable VSYNC, then the difference between LightBoost=10% versus LightBoost=100% is almost diminished by non-synchronization of fps vs Hz, although it becomes noticeable again in source engine games if you really uncap your framerate (e.g. fps_max 999) to eliminate unexpected harmonics between fps and Hz. (fps_max 250 = harmonic of 10 Hz = 10 micro stutters per second). Even a GTX 680 can run at 250fps inside closed spaces in a source engine game, so it's very easy to get unexpected stutters caused by harmonics between fps and Hz. For motion-blur-sensitive people, stutters produce a very tiny amount of motion blur, which can cancel out the difference between LightBoost=10% versus LightBoost=100%.
.
Click to expand...

Would using fps_max 241 eliminate such stuttering (as it would be refresh rate * 2) or would fps_max 0 result in less stutter?
 
FlipBack said:
Would using fps_max 241 eliminate such stuttering (as it would be refresh rate * 2) or would fps_max 0 result in less stutter?
Click to expand...
If your GPU caps out at fps_max, you can get a stutter harmonic that's equivalent to (fps modulo Hz). So 241 modulo 120 equals a harmonic of 1 stutters per second (and an evil "scrolling tearline" effect that rolls once per second) during moments that your game engine caps out at 241fps.

Ideally, for minimum visible stutters during VSYNC OFF operation you want a random number whose modulo is closer to being roughly between refresh rates, but not exactly (e.g. fps_max 287). Avoiding exact numbers like fps_max 300 is good because it avoids stationary or rolling tearing effects (which can sometimes be seen at fps_max 119 or 120 or 121 ..... and for 239 or 240 or 241 -- this can lead to a slowly-moving tearline effect whenever the framerate caps out). Tearing is still existent but more faintly at 1/2 Hz harmonics (e.g. fps_max 180, fps_max 300) if you run a Titan (e.g. perfect cap-out at 300fps in Counterstrike) and your eyes are unusually sensitive to tearing. Some people swear by a full uncapping of the framerate (e.g. fps_max 999) but that could starve your engine of other resources, e.g. increasing input lag. You prefer consistent frame rendertimes and really high framerate, to get MANY tiny-offset tearlines occuring at random parts of the screen. If you are sensitive to tearing, this is far better than fewer big-offset tearlines (at lower fps_max or during moments of bad stutters). Limiting your pre-rendered framebuffers helps a lot, when uncapping your framerate.

It's very game-engine specific, but older source engine games easily cap out at 120fps and beyond (even 240fps) on a very fast system, so it's very easy to reproduce harmonic stutters in source engine games with VSYNC OFF (stutter beat frequency effect between framerate and Hz).

Human eyes can't see individual frames at these insane framerates, but human eyes are very sensitive to stutters, including stutters caused by harmonic beat frequency between fps vs Hz. It's amazing to see a big fluidity differences keep occuring when you experiment with fps_max values well beyond your Hz.

If you are not sensitive to tearing, by all means, run fps_max 121 and 241. If your GPU is slow enough that it runs at random framerates of 150-200fps and never caps out at 241fps, then fps_max 241 is not harmful looking in tearing artifacts. But if you've got a Titan and you cap out at fps_max, there's an ugly rolling tearline effect at fps_max 241 (harmonic frequency of 1 Hz). Make your (fps modulo Hz) a nicer bigger value, and a value that produces random tearline positions rather than stationary tearline positions.

Personally, if your CPU and GPU isn't going to cause input lag at uncapped framerates, just uncap framerate and limit your prerendered frames; it produces fluidity that almost looks like VSYNC ON -- but without the input lag of VSYNC ON. Or if you don't want to fully uncap, use an odd in-between value (far away from harmonics) like fps_max 287 or fps_max 411 to be safe, so you don't have ugly harmonic stutter effects whenever framerates hit the full cap.
 
Last edited:
I've tried this monitor now alongside the asus VG248QE and my S27A950D, and in the end decided to return the benq and retire the samsung and use 3x VG248QE.
The asus and benq simply refused to run alongside eachother in a dvi splitter, so couldn't test input lag.
The samsung and asus had exactly the same input lag on the millisecond almost every frame (samsung 2d vs asus lightboost), but the samsung added roughly 10-30ms in 3d mode (it varied greatly between the frames).

The XL2720T had:
*Worst uniformity. It was still not terribly bad, and a sample of trying one monitor doesn't prove anything. Noticable bleeding in top left/bottom left corners.
*Slightly worse lightboost mode than the asus, but better than the samsung (the samsung is great in a pitch black room though).
*Brightest and best standard colors in lightboost mode. (subjectively, I didn't measure it using any equipment but my eyes). This was the only thing the benq was best at.
*About equal blur to the S27A950D with lightboost off. The asus has clearly less than both even in non-lightboost mode.
*Horrible touch menu buttons that registers one click in five.
*Less apparent TN-gammashift than the asus, despite being larger. This was a nice surprise. The samsung also has less colorshift than the asus, despite being larger.

144hz wasn't listed as an allowed mode by default, and i didn't try to set it manually.
 
Last edited:
I found the touch buttons easy to use once I figured them out. Using my thumb had a much higher success percent and if I used the flat part of my thumb instead of the tip (like you would with a thumb fingerprint) it worked 100% of the time.
 
Mechanical buttons are always better than touch IMO. You can do things quicker with the Asus QE than with the BenQ.
 
I never used the buttons. I used the scroll device unit. Never had issues making changes on it with it.
 
martinje said:
I never used the buttons. I used the scroll device unit. Never had issues making changes on it with it.
Click to expand...

Agreed, why would you use the panel buttons when the monitor comes with a great controller?

I've had the XL2720T for 3 weeks now, and I'm keeping it. I've read everything I could find about 120Hz gaming monitors and I'm satisfied because I like 27" and I'm not going back to a smaller size.

Basically, if you like 24", or motion blur is the most important thing to you, get the Asus QE. If you want to game at 27", the BenQ is the best option for now. For FPS games, it's still the best monitor I've used and really has little blur compared to 99% of LCD monitors.

Also, there's different types of consumers. Some consumers are really into testing components and motion blur reduction -- for them, the QE is the best option. For me, I'm looking for the best experience in FPS gaming (BF3). For this purpose, I want a 27" 1080p screen so I can set the highest field of view while having enemies remain large in size. And I like the black equalizer for games as well.

The fluidity in-game is wonderful. I've been averaging about a 5 K/D the last few weeks. My biggest issue now is that I've come to expect the fluidity of 120 FPS and can tell when my framerate drops below 100 or so. When it drops to 60, it looks downright laggy now, so I've really become spoiled. Even with 2x 660's in SLI at medium settings, it will drop below 100, depending on the map (multiplayer). So I might have to change cards again when the 7-series rolls out.
 
FoolishBose said:
I'm itching to buy this monitor but was waiting on more reviews to pop up. I'm a little worried by the disclaimer on PC Monitors. I don't want a run of the monitor that has 'issues'.
Click to expand...

PC Monitors have updated the disclaimer explaining that the problems only occurred on UK test samples and not on retail stock.
 
I went to Fry's last night and they had an Xl2024T there. I was informed that they have quit carrying it at Fry's.


They let me play with it there.....It was in the middle of a bunch of monitors on display.

I could not get the basic picture quality even close to any of the basic 24 and 27 inch monitors ( TN and IPS) around it. Was very disappointing.
 
I saw this at Fry's....ASUS VS278Q-P

Yea I know it isn't light boost..but it did have a great basic picture
 
BigBrutal64 said:
But, what some of us need is these questions answered:

What is the best Light Boost monitor out there right now?

Is there a better 27 inch over other 27 inchers---Benq vs Asus?

Is 24 inch Asus better than the 24 inch Benq?


Does anyone have a price for the ASUS VG278HR?
Click to expand...

Of all the models available, the BenQ 24 incher seems to be regarded as the best.
 
BigBrutal64 said:
I could not get the basic picture quality even close to any of the basic 24 and 27 inch monitors ( TN and IPS) around it. Was very disappointing.
Click to expand...

Just got this monitor yesterday, as everyone has said, looks horrible out of the box. As I've tweaked it further though it's gotten better and better. I'm almost very happy with it but I'll probably require a few more days of tweaking.
 
Are the viewing angles really that bad? I don't expect IPS quality, but is it at least viewable by 2 or 3 people besides the user with minimal color loss?

Take a look at this review image from the VG278HE:

http://www.tftcentral.co.uk/reviews/content/asus_vg278he.htm#viewing

I wonder if anyone knows of a similar image for the Benq XL2720T? Or if it would be possible to make your own if you are willing to help potential buyers.
 
CrisisCauser said:
Are the viewing angles really that bad? I don't expect IPS quality, but is it at least viewable by 2 or 3 people besides the user with minimal color loss?

Take a look at this review image from the VG278HE:

http://www.tftcentral.co.uk/reviews/content/asus_vg278he.htm#viewing

I wonder if anyone knows of a similar image for the Benq XL2720T? Or if it would be possible to make your own if you are willing to help potential buyers.
Click to expand...

Probably the exact same panel used for both monitors. I have the xl2720t and angles are about what you'd expect from a large TN panel.
 
could anyone tell me if I could run an older game at 800x600 @ 120hz with this monitor ?
 
thanks!

last question

54650.png


is this anything to be concerned about ? I currently have a 60hz samsung 1ms monitor, so I wouldn't want the input lag/delay (no idea what the difference is) to be higher than it is now...
 
shedokan said:
huh? BenQ 144Hz higher input lag than all of these laggy monitors? I'm confused
Click to expand...
That's an input lag measurement at 60 Hz, through HDMI port, using the Leo Bodnar input lag tester.
It's not valid for 120Hz.
______________________

Some good news, I resumed work on my Arduino Input Lag Tester.
I have invented the BlurBustersSCOPE:

inputlagmeter.png


This is the old screenshot. The software is greatly improved now, and I've finally fixed bugs degrading accuracy. The input lag results are now repeatable (0.5ms accurate, and in some cases 0.1ms accurate) It is running a flashing square over Direct3D, in the very center of the screen, just like the Leo Bodnar input lag tester. ("Cable Lag" = USB cable between Arduino and computer. It has to compensate for this cable lag, to get accurate measurements).

It features:
-- Direct3D for the fastest possible delivery of graphics to monitor
-- Automatic compensation for USB cable lag (subtracts USB cable lag)
-- Automatic compensation for Windows compositing buffering (subtracts a vsync cycle)
-- Automatic detection of when the sensor is on the screen
-- Works at any refresh rate
-- Open source hardware. Costs only $40 to build. Can use almost any Arduino from www.arduino.cc
-- Easier to build than a computer. No soldering required (just adding two components to a prebuilt Arduino board!)

I am expanding it to include three flashing squares -- top, center, and bottom -- and force full screen mode, so that the flashing square is at "perfect" locations.

It's ready for beta testing, I just have to post build instructions.
Please send me a PM if you want to build this input lag tester (it's easier than building a computer -- no soldering needed!) It consists of an Arduino from www.arduino.cc then you buy a photodiode & a 100Kohm resistor, and a few wires. You simply connect two components to the Arduino, and viola!
I finally fixed the software to eliminate errors.

For those asking for LightBoost input lag measurements; that will be published as a new Blur Busters Blog entry -- keep tuned.

P.S. Contact me if you want to build and test an Arduino Input Lag Tester (Beta)
 
Last edited:
yeah but even at 60hz it's about the same as an IPS panel.. can 120hz improve it THAT much ?
 
Those numbers are for 60Hz + HDMI.
cpaqf1 said:
thanks!

last question

54650.png


is this anything to be concerned about ? I currently have a 60hz samsung 1ms monitor, so I wouldn't want the input lag/delay (no idea what the difference is) to be higher than it is now...
Click to expand...
Those numbers are for 60Hz + HDMI.

This is NOT VALID when running 120 Hz, and through the DVI port.
 
cpaqf1 said:
yeah but even at 60hz it's about the same as an IPS panel.. can 120hz improve it THAT much ?
Click to expand...
If someone wants to send/loan me a BENQ XL2720T to test with the Arduino Input Lag Tester, or someone is located in Toronto area, I'd be happy to measure the XL2720T lag.

Alternatively, if someone with Arduino knowledge and the knowledge how to compile a program, I can write up instructions on how to build the BlurBustersSCOPE so you can do the lag measurements yourself.

There are many sources of inaccuracy in input lag measurements
-- Blanking interval length
-- When to start measuring -- at graphics output? at monitor input? Bfore or after HDMI encode/decode? Some stages are easier to measure input lag, others are harder to measure.
-- Some ports can have a full frame less lag
-- VGA vs DVI vs HDMI
-- Measurement of top edge vs center vs bottom edge (can vary by ~16ms from top edge to bottom edge due to scanout)
-- PWM even affects lag. Brightness=0% has approximately 0.5ms more input lag than Brightness=100% due to the black gaps between flickers.
-- Pixel transition delay but the human can see the partial transition early. (e.g. a slow 10ms pixel transition may be mostly visible after only about 3ms) -- does an input lag tester stop measuring upon complete transition, or upon partial pixel transition? That can be a challenging question to answer.
-- etc.

To understand the whole input lag chain at a consumer level, see Anandtech Exploring Input Lag Inside and Out.
It does not cover this to the sub-millisecond detail (e.g. displayport packetization, port transmitter electronics, Auto vs Reduced blanking interval, slow pixel rise curve vs instant lightboost flash, PWM effect on input lag, etc) but it is an excellent way to understand the whole input lag chain at the multiple-millisecond level.

longlag.png


-versus-

bestcase.png
 
Last edited:
I see, would it be reasonable to assume that when running at 120hz, we could split the delay in two, so around 11 ?
 
NCX said:
PRAD has oscilloscope measurements, the BenQ has negligible lag.

5.8ms signal delay @60hz

3.5ms signal delay @120hz

http://www.prad.de/new/monitore/test/2013/test-benq-xl2720t-teil6.html#Bildaufbauzeit
Click to expand...

thanks, I bought the review to take a closer look, I'm trying to compare with the asus 27 I have here, Asus VE278Q but they don't have the same kind of review for it. Do you think the input lag is equal or less ?

I can't really compare both reviews for the BenQ either cause one gives 22ms @ 60hz and the other gives 7-8ms..
 
Last edited:
You must log in or register to reply here.
Back
Top