[GUIDE] to the problems with the ASUS ROG SWIFT - Overclock.net - An Overclocking Community

Forum Jump: 

[GUIDE] to the problems with the ASUS ROG SWIFT

Reply
 
Thread Tools
post #1 of 48 (permalink) Old 11-26-2014, 04:41 PM - Thread Starter
New to Overclock.net
 
Mountainlifter's Avatar
 
Join Date: Nov 2014
Posts: 191
Rep: 26 (Unique: 22)
Gross Issues:

What follows is an exhaustive list of problems which users have reported with the Swift.

G1.Dead or Bright/Stuck pixels or Bright Spots:
http://youtu.be/q2ProvWa6h4
http://youtu.be/PFTpnAOkncI
http://youtu.be/B5EZkLCPJ2g

G2.Excessive Back-light Bleeding or Clouding:
Clouding seems happen within one month of use.
source

G3.Blurry text or flickering text:
http://youtu.be/Umx7ZvDKzdg
http://youtu.be/BcUhVYptfTk

G4.Flashing set of noisy stripes:
http://i.imgur.com/bWX21EP.jpg
Source

G5.Persistent horizontal and vertical scan/grid lines:
http://i.imgur.com/cIRszXp.jpg
Source
http://s168.photobucket.com/user/popdude338/media/IMAG0042.jpg.html & http://s168.photobucket.com/user/popdude338/media/IMAG0033.jpg.html

G6.Backlight flickering with/without gsync:
http://youtu.be/QV7yHw1hzyc
Source

G3 and G4 or G5 seem to happen together. G3 and G6 also seem to happen together. And these problems also appear after a major crash in BF4.

G7.ULMB backlight banding:
http://youtu.be/z8a09vjGFoU

G8.Twitching?:
http://youtu.be/aElSxcE8KV4

G9.Slow green pixels?:
http://youtu.be/wvn9F_IA6zs

G10.Won't work above 60Hz:
http://youtu.be/zTODqFO6iY4

G11.Total breakdown on startup or after a few days:
http://youtu.be/1AGi4BFOOBA
http://rog.asus.com/forum/showthread.php?51701-PG278Q-Half-green-screen-on-first-use

G12.Flickering On and Off :
http://youtu.be/ST7FPR6JF9g
Source

This list is not complete. But these are the gross (hence, 'G') issues that have been reported. Solutions vary the spectrum from replacing the DP cable to a full RMA. (I have linked the youtube videos of various persons without their permission with the assumption that they would be willing to share their monitor experiences.)
________________________________________________________________________________

Introduction to three other problems:

If you do not have any of the above listed issues then you are lucky. (So am I). But then these problems have developed over time spans of three days to three weeks to three months. So, be on the lookout.

On this thread, I mainly wanted to discuss three other issues that mostly seem to be going unnoticed or are not well documented. So, welcome to a discussion of the subtler problems of the Swift. I'm calling them the three Subtle problems; 'subtle' only because they are tougher to detect:

  • S1. Green Blur and Negative Ghost
  • S2. Alternate bright and dark parallel lines of pixels
  • S3. Inversion test behaviour

Especially S2 goes unnoticed by most users (though there is a hefty thread on it over at the ROG forums) who then decide that their monitors are in perfect condition (as did I). S1 seems to be rare but I won the lottery on that. And S3 is confounding because it is inconclusive but more on that later. S3 is also believed to be the cause of S2 and may be the cause of S1 too. But in my discussion, I am keeping all three separate.

The aims of this thread are:
  • A1. to let new owners get a handle on what problems the Swift might exhibit, and/or help the public make an informed purchase decision. (The intention is not to scare anyone away. Though the problem list is huge, the price and popularity of the monitor, the number of units sold, the fact that it implements a whole lot of new tech may be amplifying how we perceive the real extent of it's QC failures). Therefore, this is not a smear campaign of any sort. I still like Asus products and most of my rig is made of their stuff.
  • A2. To create a few standardized tests that allow existing owners to see if they have the above listed subtle problems.
  • A3. To see if there is a correlation between newer manufacturing dates and the absence of these issues. I certainly hope that the monitor is not working "as intended" with these three "S" problems.
  • A4. Finally, if all the reported problems are listed or documented in one spot, perhaps we might get Asus' attention. A fix to these subtle problems, hardware or software, if possible, would be fantastic.

A final reason why I wanted to start a new thread instead of putting this wall of text into the owner's club is that, in that thread, a discussion of these problems is often interspersed with discussions about availability, RMA experience etc. and reading or searching through 300 pages looking for some particular thing will only become increasingly tedious.

________________________________________________________________________________

S1: Green Blur and Negative Ghost

When a fast moving object leaves behind an ephemeral remnant over the pixels it had travelled over, we call that a ghost. It is different from blur.

We all know what blur is and blur itself can be annoying. To overcome or avoid blur, pixels are usually forced to respond faster from one refresh to the next when changing to a different color. But over driving the pixels in this way causes ghosting because over-driving introduces an overshoot in the step response of the pixel. "Overshoot" means a pixel responds faster to the command to change it's color but overshoots that color a bit before settling on the correct color.

TESTS1_A:

To see the effect of overdrive for yourself use the blurbusters ghosting test http://www.testufo.com/#test=ghosting At the three overdrive settings available on the swift, this is what you should see behind the moving UFOs:

OD Off: blur
OD Normal: very little blur, no ghost
OD Extreme: a ghost that looks like the negative of an photograph. This is what I'm referring to as a negative ghost.

(Note: For the sake of standardization of these tests, leave the brightness at 50 and contrast at 70, color at "user mode" with R, G and B at 100, the gamma in the NV control panel at 1, and frequency at 120hz.)


But this negative ghost is not necessarily a "problem". It is simply the way the panel behaves when the pixels are being driven faster.

Another way that blur is avoided in the swift in particular is with the ULMB feature. The backlight is strobed in sync with an image on the screen so that the pixels are lit very briefly and this eliminates blur. But what I observed on the blurbusters ghosting test is that ULMB produces a negative ghost that is worse than or just as bad as what is seen with the extreme overdrive setting.

ULMB: crisp negative ghost

(Note that ULMB works only at refresh rates of 85hz, 100hz and 120hz. The monitor automatically grays out the ULMB option in the OSD when it's run at 144Hz refresh. Similarly, when ULMB is ON the overdrive settings are grayed out.)

Neither the extreme overdrive's nor the ULMB's negative ghost are apparent or bothersome if you are playing games that look like battlefield 3 or far cry 3; worst-case, you might see outlines of objects briefly as if you are playing borderlands. However, they can be a big issue when playing brightly colored 2d games like hearthstone. Furthermore, you will notice the negative image as a conspicuous trail when you move windows around on the desktop.

TESTS1_B:

While the negative ghost is normally seen on overdriven panels, the swift's problem is that a transition from white to grey/black (or the reverse transition) leaves a green residue or blur. And it looks more like a blur than a crisply outlined ghost.

If you perform the blurbusters inversion artifacts test (moving solid colors thick) at a speed of 11 13 or 15 pixels/sec, you may see a green trail at the leading and trailing edges of the gray bar. If you have a very bad case of the green ghost, you may even see it at the trailing edge of the black bar.

This is what you might observe at the different settings:

OD off: Lot of green blur
OD Normal: reduced green blur
OD extreme: crisp dark green rectangle at the leading edge of the gray bar.
ULMB: same as with OD extreme.

(Note: For the sake of standardization of these tests, leave the brightness at 50 and contrast at 70, color at "user mode" with R, G and B at 100, the gamma in the NV control panel at 1, and frequency at 120hz.)

Unlike the negative ghost, the green blur (especially with OD off) is very visible in games especially battlefield 4 and can be bothersome. Moreover, unlike the negative ghosting, this is not "normal" behaviour for any panel, as far as I know.

TESTS1_C:

Since a transition from white to gray/black produces the green blur, I decided to perform one final test. I created a video which alternates black and white frames at 120 fps using virtual dub. Running the video in fullscreen I observed the following:

OD off: screen is green
OD Normal: screen is greyish
OD extreme: screen is white-ish
ULMB: screen is white-ish

(Note: For the sake of standardization of these tests, leave the brightness at 50 and contrast at 70, color at "user mode" with R, G and B at 100, the gamma in the NV control panel at 1, and frequency at 120hz.)

But the surprising thing was that when the video was running in fullscreen, I was reducing the contrast because the flashing video was too bright and hurting my eyes. I found that as the contrast dropped one notch below 50, that is, from 50 to 49, the green color vanished (at OD off setting, ofcourse.). That is, just to be clear, green at 50 and above, no-green at 49 and below.

If you’d like to try this out for yourself, you can download the video here: https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/BW_flashing.avi

But be warned: If you are susceptible to seizures, DO NOT watch this video. Even if you are not susceptible to seizures, looking at this flashing video for a long time can cause headaches! (It did for me.)

SOLUTIONS TO AVOID GREEN BLUR IN DAY TO DAY USE:

Solutions to overcome green blur includes one of the following:
1. Reducing the contrast to 49 and below will remove the green blur on TESTS1_C (OD off) but it will still appear faintly on TESTS2_B but you will no longer see its effects in games.
2. Leaving the overdrive at extreme or using ULMB always will also get rid of the green blur in TEST_S1B and TEST_S1C and also in games.

POSSIBLE CAUSES FOR GREEN BLUR:

1. The green pixels are slower to respond than the red and blue pixels. I am not a panel expert but if you take a look at problem G9, this explanation may sound a little less ridiculous.
2. This could also be due to incorrect VCOM voltage calibration for green pixels, aka the pixel inversion problem. More on this inversion stuff under S3.

OTHER COMMENTS:

Moire Patterns: Moving this image below slowly from left to right should produce what seems like a colored pattern.


You can also take a look at this youtube video for a demonstration of the effect. (credit: DeadEye)

But if you put your nose to the monitor you will see that only two colors are produced - green and magenta. So, this is simply a repeat of what happens with the blurbusters moving solid colors (thick) test. Namely, a transition from white to black produces a green trail. However, I do not know why there is a magenta trail also in this test.

________________________________________________________________________________

S2: Alternate Bright and Dark Vertical Lines of Pixels

If you’ve read anything on discussion threads about the swift, you’ve probably heard about the “inversion thing” or “screendoor effect” or “inversion lines” or “incorrect pixelmapping”. I call it “alternate bright and dark vertical pixel lines” because that is what it is. This problem is different from the persistent vertical or horizontal lines (see G5) that appear even on your desktop along with the blurry text problem. It is also different from the weird streaks you will see if you perform the lagom inversion tests (which we will discuss in the next section). No, this problem goes totally unnoticed because it is ephemeral and hard to detect. You might see it in the corner of your eye and think to yourself, as I did when playing far cry 3, that it was your imagination.

Like many others who purchased this monitor, I came home and took careful pictures of the pixels on the monitor to determine if I had this problem or not. Pictures like these:


original R (full size): https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/R.jpg
original G (full size):https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/G.jpg
original B (full size):https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/B.jpg


And then I went to the owner’s club and declared that there were no “vertical lines” with my monitor. The problem is that it does not appear on the desktop (unless you force some distortions using the lagom/blurbusters inversion tests) but appears only in the motion of bright fast moving objects in videos and games. So, a static image will not help you see the problem.

As for how this issue looks exactly, the two best videos and the best image I could find for this issue are:

If you still don't see what is going on, I’m going to give two tests for you to see for yourself that your monitor has this problem. Because no one can be told what the “Alternate Bright and Dark Vertical Lines of Pixels” problem is. You’ll have to see it for yourself. But be warned that once you know it’s there, you will start seeing it everywhere; on the 3d spot orange triangle in bf4, in hearthstone, in every muzzle flash in fps games etc. It might ruin your enjoyment of the monitor. If you'd rather be ignorant and happy, I do not blame you. tongue.gif

TESTS2_A: (red/orange color)

Download this video https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/farcry3_red_test_compressed.avi (I know the size is large but I couldn’t compress it further without compromising the test.)

It’s an 80 fps video with alternating frames – one frame with muzzle flash from far cry 3 and the other is the same scene without the muzzle flash. Using a video to check for this issue eliminates the need to hunt for it by launching games.

You have to run the video in windows media player with fraps to ensure that you are running at 80fps. The muzzle flash will have three distinct regions – yellow, orange and red. While the muzzle flash is flashing on and off, look to the red portion with your face about 1 foot from the screen.

What are you looking for? THIS:

zoomed in


Full sized images here:
https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/Seen1.jpg
https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/dell%20FC3.jpg


TESTS2_B: (green color)
Download this video https://dl.dropboxusercontent.com/u/2371050/ROG%20SWIFT/Hearthstone_test_compressed.avi

Repeat TESTS2_A. This test demonstrates the same effect with bright green colors. Look at the bright green edge of the flashing card. If you’re still really sure you do not have this problem, please let us know.

(Note: For the sake of standardization, leave the brightness at 50 and contrast at 70, color at "user mode" with R, G and B at 100, the gamma in the NV control panel at 1, and frequency at 120hz. Also, put the OD setting to extreme or use ULMB mode for TESTS2_A and TESTS2_B. Otherwise, you might find it tough to see the lines though users have reported seeing it at any OD setting)

POSSIBLE CAUSES/SOLUTIONS FOR ALTERNATE BRIGHT AND DARK VERTICAL LINES OF PIXELS:
  1. A problem with the panel itself aka the pixel inversion voltage problem. More on this inversion stuff under S3.
  2. Faulty Nvidia G-sync module. It has been stated by an asus rep on the ROG forums that there is no firmware that can be applied to the swift. This is because the scaler electronics found on regular monitors has been replaced with the g-sync module. If you assume the AOC panel, the GPU and the DP cable are working fine, then the problem could the gsync module itself.
  3. People have tried replacing the DP cable but that has not solved it.
  4. GPU Drivers? Many driver sets have come and gone since the SWIFT released but we are yet to see a resolution of this issue. I’ve run TESTS2_A and TESTS2_B after clean uninstalling the Nvidia drivers and I can see the effect then too.
  5. Over at techreport's review, Scott Wason mentions that he sees this problem and speculates that it could be due to FRC dithering. Source
  6. TFT central's review doesn't explicitly mention this problem but take a look at this capture of negative ghosting below. The blur/ghost on the benq monitor doesn't have alternating lines.


________________________________________________________________________________

S3: Inversion Tests

couldn't get around to this before I sold my rog swift unfortunately. Pcper has done some good work investigating this and other panel issues http://www.pcper.com/reviews/Displays/Acer-XB270HU-27-1440P-144Hz-IPS-G-Sync-Monitor-Review

________________________________________________________________________________

Let's be scientific and work together

Reply to this thread in this way:

Manufacturing date: September 2014

TESTS1_A: yes, I see the negative ghost.
TESTS1_B: yes, I see the green blur. (No, I'm not color blind)
TESTS1_C: yes, the flashing video is not green at 49 and below. (Or I didn't do the test cause I didn't want a headache).

TESTS2_A: yes, I have the alternate bright and dark lines on the FC3 video.
TESTS2_B: yes, I have the alternate bright and dark lines on the Hearthstone video.

GPU: GTX TITAN SLi
DP: supplied cable or bought separately.

Comments/discussions/test-suggestions:
  • What egregious problems from the G list are you having? how did you solve it?
  • I hope this thread helps, you guys! Credit goes to the people who made the youtube videos, took pictures and shared their findings.
  • If you want me to add more items to the list of problems, give me a name and an image or video.

My Rig:Warning: Spoiler! (Click to show)
CPU: i7 3770K OC 4.5GHz @ 1.20V
MoBo: Asus Maximus V Formula Z77
RAM: 4*8GB Corsair Vengeance Pro 2133MHz
GPU: Asus Strix 1080Ti
SSD: Samsung 960 EVO M.2 NVMe
KB/MS: Corsair Vengeance K65, SteelSeries Rival 310
CPU Cooler: Noctua NH-D14
PSU: Corsair AX1500i Titanium PSU
Sound Card: Asus Xonar Essence STX
Headset: Beyerdynamic T90
Monitor: Acer Predator X34
OS: Windows 10
Mountainlifter is offline  
Sponsored Links
Advertisement
 
post #2 of 48 (permalink) Old 11-26-2014, 05:10 PM
New to Overclock.net
 
Smokey the Bear's Avatar
 
Join Date: Jun 2011
Location: Canada
Posts: 1,390
Rep: 61 (Unique: 51)
Manufacturing date: October 2014

TESTS1_A: No ghost on normal.
TESTS1_B: No, I see no green blur.
TESTS1_C: Yes, the flashing video is not green at 49 and below. Was not green for me at 50.

TESTS2_A: No, I don't have the alternate bright and dark lines on the FC3 video.
TESTS2_B: I had difficulty playing this video smoothly, but noticed nothing.

GPU: GTX 980
DP: Supplied cable

I also get moire patterns and terribly uneven backlight.

Smokey the Bear is offline  
post #3 of 48 (permalink) Old 11-26-2014, 05:33 PM
Linux Lobbyist
 
Klompet's Avatar
 
Join Date: Aug 2014
Posts: 116
Rep: 6 (Unique: 6)
Great thread OP, should be stickied imo. That said I think I've been convinced not to order a swift after seeing it all compiled - and I was about a week away from ordering!
Klompet is offline  
Sponsored Links
Advertisement
 
post #4 of 48 (permalink) Old 11-27-2014, 05:46 AM
Linux Lobbyist
 
Apocalypse Maow's Avatar
 
Join Date: Oct 2011
Posts: 610
Rep: 15 (Unique: 13)
Thanks for posting all this in one spot! +rep
Apocalypse Maow is offline  
post #5 of 48 (permalink) Old 11-27-2014, 08:14 AM
New to Overclock.net
 
KraxKill's Avatar
 
Join Date: Aug 2014
Posts: 529
Rep: 47 (Unique: 34)
First of all. Mad props to the OP for taking the time and effort to compile this.

I have been taking note of all the issues, but having them all listed in one place is huge. I have already sent back two ASUS Rog Swifts via RMA (July build date) when my RMA was returned damaged and am currently waiting on a replacement with an October build date. Will provide Test 1A, 1B, 1C. Results when new unit arrives.

My original issue besides the vertical striping (lines) which were annoying but IMO less bothersome to me than the issue which prompted me to send the unit(s) in. The blurry text which made the display unuasble. The best way I can describe the blurry text problem is that it gives you the impression that you can't focus and need your eyes checked.

On my screen the far right hand side of the screen was clear as when I got it, the left, middle and middle right were blurry.

I could clearly tell that there is an issue and could consistently replicate/demonstrate it by simply right clicking on different portions of the desktop and reading text in the prompt.

It's clear on the right with great definition, the opposite on the left. Fonts are hard to read, fuzzy. Some letter like the "e" and "a" have the centers filled where as on the right hand side (Clear side) the centers a clearly visible and distinct.

When I first noticed it, I thought I needed to go see an optometrist but upon further inspection it's the display.

I have tried cables, different nvidia drivers (clean install and not), different refersh rates, connecting the display with the computer on, resetting the display to defaults etc. Nothing seems to fix the issue.

Here are some pictures to add to the thread.

Again huge thanks to the OP. Subscribed and hopefull that this thread helps ASUS fix these issues. I loved the display while it worked. The fluidity of 144hz when playing at 60fps or above with gsync is a game changer.


Left Edge

IMG_1602-1.jpg

Right Edge

IMG_1603-1.jpg
KraxKill is offline  
post #6 of 48 (permalink) Old 11-27-2014, 09:18 AM
New to Overclock.net
 
Deadeye's Avatar
 
Join Date: Aug 2012
Posts: 350
Rep: 3 (Unique: 3)
Nice one! About my Moire i personaly dont think that its like green ghosting. These pixels dont act like ghosting. I will upload later that green ghosting issue that i had and only thing that fixed it is lower green color to 49 or more i have i think 48.

P.s. It's Deadeye smile.gif

Deadeye is offline  
post #7 of 48 (permalink) Old 11-27-2014, 11:01 AM
New to Overclock.net
 
Dmitriy's Avatar
 
Join Date: Jun 2013
Location: Boston MA
Posts: 945
Rep: 56 (Unique: 51)
Thanks for your work. Also you don't need someone's permission to put a link to their video here. Internet is all about links.

Dmitriy is offline  
post #8 of 48 (permalink) Old 11-27-2014, 11:08 AM
Meow!
 
djriful's Avatar
 
Join Date: Aug 2011
Location: Toronto, Canada
Posts: 8,317
Rep: 242 (Unique: 204)

So this monitor has so many problems?



~ Nothing to see here ~


djriful is offline  
post #9 of 48 (permalink) Old 11-27-2014, 11:22 AM
*cough* Stock *cough*
 
Nukemaster's Avatar
 
Join Date: Oct 2013
Posts: 2,696
Rep: 129 (Unique: 101)
I always thought inverted/negative ghosting was limited to VA based panels with excessive overdrive(like my 245t).

Learn something new every day.

Gaming
(15 items)
Media Center
(17 items)
CPU
2600K @ 4.4
Motherboard
Maximus IV Gene-z
GPU
Asus Dual 1070 OC with Artic Mono Plus
RAM
G.SKILL F3-1866C9D-16GXM Ripjaws X 16GB
Hard Drive
M4 256
Hard Drive
MX 300
Optical Drive
Liteon Slot load
Power Supply
Corsair 850HX
Cooling
H80i
Case
Silverstone FT03
Operating System
Windows 7 HP 64 bit
Monitor
BenQ XL2420Z
Keyboard
Logitech G810 Orion
Mouse
Logitech G502
Audio
X-fi Titanium Fatality
CPU
i5 750
Motherboard
h55n usb3
GPU
msi gtx 650 ti
RAM
Corsair XMS3 CMX8GX3M2A1600C9
Hard Drive
WDC Red 3tb
Hard Drive
M4 128
Hard Drive
WDC scorpio Blue 2.5
Optical Drive
Sony AD-7690H-01
Power Supply
FSP SFX 300 watt
Cooling
Geminii S
Case
SG05
Operating System
Windows 7 HP 64 bit
Monitor
SyncMaster 245t
Keyboard
Logitech Wave
Mouse
Logitech G9
Audio
Sound Blaster Omni
Other
HD PVR
▲ hide details ▲
Nukemaster is offline  
post #10 of 48 (permalink) Old 11-27-2014, 01:34 PM
New to Overclock.net
 
Smokey the Bear's Avatar
 
Join Date: Jun 2011
Location: Canada
Posts: 1,390
Rep: 61 (Unique: 51)
Quote:
Originally Posted by djriful View Post

So this monitor has so many problems?

To be perfectly honest, other than the one reason I returned mine for a replacement, I couldn't notice anything in practical use, but only used it for a night. It can't be denied there are serious QC issues. I'm hoping I get a November model this time and that most of it has been fixed. Asus would be wise to do so.

Smokey the Bear is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off