If you are going to use the ASUS PQ321Q, you’re going to want DisplayPort 1.2 support. HDMI will work, but it’ll be choppy with its 30Hz refresh rate. If you have a video card with dual HDMI 1.4 outputs, you can use both of them to drive it at 60Hz if your video driver supports it. DisplayPort 1.2 allows for Multi-Stream Transport (MST) support, letting you drive two displays with a single DP cable. But why does that matter if the ASUS is your only monitor? Because to get the full 60Hz refresh rate out of it, DisplayPort needs to see it as a pair of 1920x2160 monitors that each get their own signal.

The ASUS has MST mode disabled by default. With my NVIDIA GTX 660 Ti I had to manually enable it in the monitor for it to turn on. I’ve been told that with ATI or Intel GPUs over DisplayPort 1.2 it is automatic, but I don’t have those to test with. Once enabled, it quickly went from 30 Hz to 60 Hz while staying at 3840x2160 resolution.

Since I run multiple displays like most people, this seemed to be an ideal time to test out Windows 8.1 and its ability to offer individual DPI scaling on monitors. For this test I used the ASUS PQ321Q, connected over DispayPort, and a Nixeus VUE 30 (review forthcoming) connected over DVI running at 2560x1600. With a single universal setting, you use a percentage setting for scaling in Windows 8.1. With individual control, you use a slider more like on a Retina MacBook Pro. The percentage is hidden, which I dislike. I don’t understand why we have a different way to select the scaling level if you have two monitors versus one. Perhaps it is a beta issue, but I think they should be uniform.

Moving beyond that, when I attempted to scale the PQ321Q, I had an image that was still fuzzy instead of sharp. Thankfully a driver update (as 4K MST panels are new) fixed this issue quickly. The independent display scaling in Windows 8.1 still didn’t work the way I wanted it to. The choices are unclear, including which monitor you are adjusting, and I never could get it setup exactly how I wanted it. I wound up setting it to 150% for both displays and dealing with my 27” running with larger icons than I prefer.

Now I have an effective 2560x1440 desktop, only everything is sharp. Amazingly sharp. It is like moving from my iPhone 3G to the iPhone 4 and its retina screen. The text as I write this in Word is crisp and clear, and editing gigantic spreadsheet in Excel is much easier when the cells are so easy to read. Unfortunately not every application in Windows plays well with DPI scaling.

Chrome is scaled 150% as Windows asked, but it is hazy and blurry. Disabling DPI scaling for the application and then scaling to 150% inside Chrome produces crisp, clear text. Firefox also didn’t scale automatically, but it has a setting to adjust to make it follow the Windows DPI scaling rules. Once set, Firefox looks very nice and crisp. For most people, that setting should already be set to follow DPI scaling.

Finding a chat client that works well is a challenge. Both Pidgin and Trillian don’t do DPI scaling and are fuzzy by default. Another app that had issues is Steam. Right-clicking in the System Tray icon brought up a menu in the middle of the screen, where it would be without DPI scaling. The reality is that some apps are great and support DPI scaling, and some need work, just like when the retina MacBook Pro was released. Evernote looks great, but Acrobat is a fuzzy mess. This is all a bit of growing pains, but I find myself disabling DPI scaling on applications that don’t support it because I prefer tiny and sharp to fuzzy and large.

Because the 2560x1440 resolution is what I’m used to with my usual 27” monitor, I found there to be no real difference in how I used the ASUS monitor. I typically split items to different sides of the screen, with Word on the right and Evernote on the left as I type this. The application that benefitted for me was image editing. Being able to fit more on the screen, or zoom in to higher levels, made working with images on the ASUS better than on a 27” of the same effective resolution. I don’t do that much image editing, but for the work I have done it has been wonderful.

You’ll also quickly find out how much people need to go back and fix up programs or websites to use images and text separately. Text combined in an image scales very poorly, but is often easier than doing proper layout for two separate elements. I feel a bit bad for all the developers that need to go back to fix everything to work with high-DPI screens, but that time has come.

The only way to sum up daily use of the ASUS PQ321Q is “awesome”. It’s not perfect, but much of that is the fault of Windows or other programs and websites. When you have something that can scale and look correct, it is amazing how much the extra pixel density and sharpness helps. Yes, this is the future for displays, and we are entering the transition period to get there.

Introduction, Design and Specs Internal Scaling, Brightness and Contrast
Comments Locked


View All Comments

  • DanNeely - Thursday, July 25, 2013 - link

    Probably never since margins are higher on 16:9 screens; a 21" 2560x1440 screen could be made from the same line just by cutting the panels smaller.

    In that general size bucket though I'd like to see them jump directly to 4k too though for ~200DPI. This is high enough to make AA mostly unneeded when gaming and to allow for 2:1 scaling for legacy apps.
  • jasonelmore - Wednesday, July 24, 2013 - link

    UHD or 4k is gonna be good for the living room simply because TV's will get bigger, and that's where 4k really shines.

    Back in the day, i remember 32" was massive. Then when HD first came out, 42" was the standard when talking about big screens. Now 55" is the new standard, and 70 inch will probably be common place in the next couple of years. 4k on a 70 inch will look great.
  • DanNeely - Thursday, July 25, 2013 - link

    Maybe. 70/80" TVs require rearranging your living room in ways that smaller sizes don't; and take up enough space that in smaller houses they've often impossible just because you don't have any blank wall sections that big that you can orient furniture around.
  • yhselp - Wednesday, July 24, 2013 - link

    PixPerAn Feedback

    The PixPerAn test looks great, you’ve done nothing wrong; on the contrary, the idea is to show the realistic best and worst case scenario out of which the former should be more representative for most users. However, some of your results, namely pictures 2 and 3, seem very optimistic. In my opinion, and from my experience, pictures 4 and 5 show the real best (4) and worst (5) case scenarios. How did you capture the results and under what settings? I could try to replicate the test on my U2311H and compare the results to those that I did when I purchased the monitor and upload the results.

    It might be a good idea the crop the relevant pictures (those above), pick out two that are most representative for relevant best and worst case scenarios and create a single picture instead of a gallery. If you decide that more cases should be shown that’s perfectly alright, but an organized single picture would still be easier to read. Also, some context for readers unfamiliar with what this test shows and how to read it would be quite useful.

    It’s a common misconception that monitors with higher response time than 2ms are not fast enough. Believe it or not, a lot of people, especially gamers, steer clear of monitors with higher response time. In reality things are much different; the ms rating alone says little. Not all response time ratings are created equal. One might write an article solely about the response time of a monitor, however, without going into technicalities such as RTC, PixPerAn can give an easy to read and understand representation of what a user can actually expect to experience in terms of RT in a monitor. For example, a lighter ghost image or a pale overshoot behind the moving car is usually harder to spot in actual use. PixPerAn is also good for testing stereoscopic 3D ghosting and S3D performance in general; you can clearly see the benefits of NVIDIA LightBoost in the test.

    A picture is worth a thousand words, especially in this case. Just by looking at a PixPerAn result you can derive almost everything you need to know about how fast a monitor should actually feel like (and what kind of visual artifacts you might experience). I would be really glad if you make this test standard on all your future monitor reviews, at least here on AnandTech. It wouldn’t take much space in the reviews and should be relatively quick to take. Please, let me know whether you think it’s a useful test, etc.

    Thank you for the great review.
  • sheh - Wednesday, July 24, 2013 - link

    3D graphs of response times for all combinations of start/end pixel lightnesses, which were in Xbitlabs reviews, are also interesting.
  • DesktopMan - Wednesday, July 24, 2013 - link

    Response time: 8ms GTG

    Everyone who writes about Igzo (including Chris in this article) talks about how Igzo allows for higher electron mobility which makes a faster display. Then why is the GTG this bad?

    Does it do 1080p@120hz? Might not matter if it really is this slow though.
  • ChristianLG - Wednesday, July 24, 2013 - link

    Hello. ¿Why the monitor you are reviewing isn’t a true 4k monitor? ¿ What is a true 4k monitor? Thanks.
  • cheinonen - Friday, July 26, 2013 - link

    True 4K, when it comes to digital cinema, is 4096 pixels wide, but the height can vary depending on the content. Since this is only 3840 pixels wide, it falls under the UHD heading, but everyone uses 4K anyway because it sounds good.
  • joshu zh - Wednesday, July 24, 2013 - link

    You can find 4K TV in stores now. I can see clearly the difference between 1080P and 4K. You should try it and you may change you thoughts about 4K TV. To me, the issue is you could not find enough 4K materials to watch now. I hope this situation will change in 2-3 years.
  • Sabresiberian - Wednesday, July 24, 2013 - link

    One of my "pet peeves" is the fact that the display industry goes out of its way to blur the monitor/TV line. They aren't the same, and selling them as though they were is disingenuous at best. It has given us a 16:9 standard where it doesn't belong - and now we are crashing up against the limits of standards essentially made with low-res and TV in mind.

    Putting speakers in a $3500 display is like putting a $5 tuner in a Marantz amplifier. Certainly if I have the money to pay for one of these things I'm not going to put up with crap for speakers! It is just about as useless a thing as can be done.

    I think the best thing I can say about this monitor is that finally someone has broken the ice, and I give Asus a lot of credit for that. Not saying it's bad, just saying for $3500 I'd like to have seen more polish (better OSD) and better calibrated performance. Nice job Asus, but you could have done better.

Log in

Don't have an account? Sign up now