[PCM2]

Our review of the MSI MPG 321URX is now live. This is a 240Hz 3840 x 2160 ‘4K’ UHD QD-OLED with USB-C (90W PD), KVM support and HDMI 2.1 support. VRR is supported via HDMI 2.1 VRR as well as Adaptive-Sync (for AMD FreeSync Premium Pro and Nvidia ‘G-SYNC Compatible’). The monitor delivered the usual QD-OLED performance including; impressive contrast, vibrant and exceptionally consistent colours and strong responsiveness. HDR was also pleasant to use on the monitor – overall the experience was similar to the Dell Alienware AW3225QF which I now use as my own main monitor. I really recommend either model based mainly on price or availability in your region, but there are a few differences I observed:

– I felt the Alienware was slightly better built, with a more solid ‘feel’ to it. The plastic felt more premium. But the MSI was by no means ‘badly built’ – just that they could’ve easily incorporated coated metal into the stand design which would make it look and feel more premium. People will have their own aesthetic preferences between the two, with the Alienware being the more ‘interesting’ or to some ‘fussy’ design, really. The RGB LED lighting feature of the Alienware is more interesting and customisable – you can at least appreciate it more from the front, including power LED customisation. Both can be considered ‘gimmicks’ but the MSI’s more so.

– I have no strong feelings about the 1700R curve of the Alienware. I tend to prefer curves on ultrawide models, but for 16:9 models I either dislike the curve (if it’s too steep) or as in this case I don’t mind it. It’s mild enough that I don’t find it obtrusive on the desktop or on games where my eyes frequently scan horizontally. And it just adds a slight extra feeling of depth for games and movies which some might quite like – but I stress the slight.

– The Alienware supports Dolby Vision, which I consider a bit of a bonus for when I watch movie content that supports it. But it’s not earth-shattering and the vast majority of PC game content is HDR10 only – even where Dolby Vision is used I feel many would struggle to see an appreciable difference between the two formats. It’s more noticeable for movies I feel because they’re often a bit shy to make use of the high peak brightness of the monitor under HDR10, but that isn’t the case for games.

– The Alienware supported 24Hz operation and 12-bit colour (the MSI didn’t). Both have niche use cases, but for most people will make no difference at all to their experience or simply won’t be used.

– The MSI has additional ‘OLED care’ features, though I feel the basics also covered by the Alienware (pixel shifting algorithm plus ‘cleaning’ cycles) are the most important to include. I like that the Alienware will resume normal operation after its cleaning cycle without a press of the power button, if the system is on standby. So you could just move the mouse and wake the system up as normal for example. The MSI seems to always turn off and need the power button pressed even if the cycle has run due to system standby rather than you pressing that power button in the first place (this has been fixed with new firmware).

– I’d say both models are on par with their interpolation performance, difficult to separate them based on that. Though I found the MSI just slightly better balanced with its sharpness level (both models set up how I like) for the QHD resolution. It was able to offset the softening of the interpolation with its sharpness control slightly better before introducing strong over sharpening. But it’s a small difference.

– The sRGB emulation mode of my Alienware sample was technically slightly more accurate (average DeltaE 1.02 vs 1.22 and more accurate 2.2 gamma tracking for dark shades), but the MSI’s gamma for the sRGB mode invited a bit of extra detail without a strong uplift or artificial look – some may like this. I also liked the way it had an Adobe RGB and DCI-P3 emulation setting, they both worked as they should (aside from a slight push to the green channel for the Adobe RGB mode) and are useful to have if you’re a content creator without access to a colorimeter or similar calibration device. And simply because they give a different look to game content between sRGB emulation and the full native gamut, as explored in the review.

– Both models (as is common with monitors) will use a bit more of their gamut than the developers ‘call for’ under HDR. So if the developers target sRGB or DCI-P3, you’ll get some shades (typically those nearer the edge of the target gamut) pushed out so they’re sometimes beyond the boundaries of that target gamut. That way, the output better correlates with the capabilities of the screen. The MSI pushes this a bit harder than the Alienware, so it presents some shades with a touch of extra vibrancy at the expense of accuracy. I don’t feel this was an extreme push and don’t feel it really departs too much from the original intentions of the creators, though. If you try to restrict the MSI’s gamut under HDR (for example, using the Display P3 mode) you’ll be met with badly calibrated output which will include upset PQ curves (bad EOTF tracking). Including a bad crushing of dark shades, they appear like SDR if gamma is set to ~2.8 for darker shades. So you really want to stick to ‘User’ as advised in the review and ‘Best Settings’ video.

– The MSI offers full USB-C support (the Alienware just has a 15W PD downstream port), including 90W PD plus DP Alt Mode and upstream data. It also has an additional upstream port (USB-B) with KVM support offered and an easy ‘one touch’ switching system with the menu.

In the conclusion another alternative to consider is the MSI MAG 321UPX, which is a slightly cut down version of the screen. It’s slightly cheaper with a matte black bottom bezel (possibly plastic rather than the subtle touch of metal on the MPG 321URX), no ‘Mystic Light’ RGB LEDs at the rear (no big loss) and cut down USB capability. It has a single USB-C port with DP Alt Mode and 15W PD, but no additional USB ports and hence no KVM support or ability to connect USB peripherals to the monitor.

The more controversial difference is that it doesn’t include user upgradeable firmware. Now this doesn’t mean there won’t be hardware revisions of the monitor through time, as happens with other models that lack this capability. Or that you couldn’t send the monitor to a service centre to have the firmware updated (I’m not sure this is offered, though). But it’s obviously a nice flexibility and very convenient being able to update the firmware yourself. I can’t confirm the current ‘state’ of the MAG 321UPX’s firmware as I haven’t tested it, but I’m hoping it’s in-line with the current MPG 321URX which is free from any really annoying bugs as a PC user. According to PS5 user feedback there’s a juddering issue with VRR on that system with the MSI (not reported with the Alienware), though MSI are aware of the issue and will fix it with firmware. Not sure how the MAG 321UPX would sit – whether it’s somehow free from that issue or whether they’d need to fix that with a hardware revision.

[sblantipodi]

There is an interesting analysis on this monitor on monitor unboxed.
After three months of “professional usage” monitor shows some sign of image retention, colour uniformity issues, ecc…

OLEDs are not ready for PC use IMHO, they are more oriented to TVs, consoles and very light PC use.

It’s unacceptable to see problems after three months only.

[PCM2]

Yes, I’ve been following Tim (Monitors Unboxed)’s progress. The current latest video for those interested. A few interesting points to bear in mind:

– Tim doesn’t run the cleaning cycle every 4 hours as ‘recommended’. He commonly uses the monitor for >8 hours without doing so, sometimes running into the 16 hour of cumulative usage mark before running the cleaning cycle. With this the monitor becomes more assertive and will run the cycle without allowing you to interrupt it.

– He will sometimes leave the monitor on for extended periods (not sure how long, I think he might’ve mentioned in an earlier video) if he does take a break from the screen. He does this to emulate how he’d use an LCD and that’s fair enough, but his general advice (I agree with this) for users is to allow Windows to send the screen to sleep after a few minutes of inactivity.

– Tim states that the burn-in he has experienced so far doesn’t cause noticeable issues during normal use, just with specific test patterns.

– I agree with his take that it’s likely that as he continues to use the screen as he is, it will burn-in over the coming months. And he will continue to use the monitor in this way, of course. It should be taken as a warning that you can’t use OLED screens extensively for productivity and you can’t treat them like you can an LCD.

But I also think it’s a misleading blanket statement that they “aren’t ready for PC use”. Many do and will use them for “PC use” without issue. I use and will continue to use them for “PC use”, which includes a good dose of productivity but also treating the screen more cautiously than Tim. I think leaving your screen and computer alone and having a break after a few hours (if not sooner) is perfectly reasonable and good practice for health reasons – and in doing so the monitor can run its cleaning cycles more regularly than 8-16 hours. I also break up productivity usage with gaming and video watching, though the video watching isn’t always full screen. I do quite a lot of split-screen productivity work with my OLEDs actually, but undoubtedly allowing it to run its cleaning cycles or just for the screen to turn off for a few minutes every now and then makes a difference. This sort of mixed usage is fairly typical for many PC users, but definitely not all PC users. It’s the difference between mixed usage (which is fine with current OLED technology) and usage heavily weighted towards productivity without frequent breaks (which isn’t).

[sblantipodi]

I admit that my previous statement “was too strong” but I think that is peaceful that giving attention to a monitor is not something really possible in a work environment.
It mostly depends on the work environment but if you work with other people,
it’s a bit difficult to say them, ehm, wait, my monitor is doing a cleaning cycle.

Imagine in putting an OLED into an office, this will not work, that’s why my previous statement.
Cleaning cycles helps with burn in but does not solves the problem with static contents during the PC usage.

I think that many people are underestimating the compromises that must be made to use an OLED in a professional or work environment,
brightness is another key factor when considering burn in…

It would be interesting to see what happens if you use the monitor at 160/200 cd/m2 along with all the “needed attentions”.

Most companies are giving two years warrenty or more because they know that most people uses the monitor for gaming but this does not mean that you are safe for two years if you use it “for PC use without extra care for the panel”.

[PCM2]

Yes, I agree with that. OLED models aren’t the workhorses LCDs are and you need to be mindful with their usage in a way that won’t be practical in office-style work environments.

[Lockjaw333]

Hi Adam. Big fan of your work. Just wanted to give my two cents on both of these monitors. I’ve owned the MSI mpg 321urx for about two months, and I just picked up the AW3225qf a few days ago. Interestingly, my findings agree in some spots, and seem to differ a bit with what you described in the original post in others.

I want to preface these thoughts by saying that I have about a decade or more of experience with using colorimeters to profile and calibrate my displays. I’d consider myself fairly sensitive to inaccuracies or other PQ oddities. I feel like at this point, I have a decent eye for what’s “correct”, or close to developers intent.

Curve vs Flat: First to discuss the curve, I found it to mostly be unnoticeable during normal use. If anything, I think its a bit of a welcome thing, as at 32″ the edges of the screen can seem a bit far away on the flat MSI panel. The subtle curve on the AW draws them in just a tad. I found it a bit easier to use in terms of eye tracking for fast moving content when gaming. Having said that, I did get an odd sensation at times that the panel was crooked or not level, even though it was indeed at a level position. Most noticeable during cut scenes in games with black bars at top and bottom. I think the straight lines of the black bars gave me a weird optical illusion against the bottom curve of the panel, causing this weird sensation of the monitor being crooked. Again it was subtle and not a deal breaker by any means. Overall I found the curve, at least for gaming, to give a bit extra immersion, and to be mostly unnoticeable during use.

Gamma: The big difference I saw between these two panels is in apparent gamma of the image, both in SDR and HDR. As a general statement, the MSI feels darker (higher gamma) in both SDR and HDR, while the AW is significantly brighter (low gamma). This is immediately apparent in scenes with dark and shadow detail.

The AW near-dark gamma is too bright (low gamma), causing near black and shadow detail to be lifted way too high. This is apparent in both SDR (creator- gamma 2.2), as well as in HDR. For SDR creator mode, it seems they calibrated to sRGB gamma curve instead of 2.2, which as you know has a low gamma as you approach black- near black tones are closer to gamma 1.8. On the LAGOM black level test for example, the first row of blocks are significantly lifted against the background. The MSI in SDR (sRGB mode), in contrast, has a much darker first row of blocks, and as is typical of a panel that tracks 2.2 closely, the first few blocks are barely visible and almost seem to blend with the background. In practice in SDR content, this gives the MSI a more “correct” look IMO. I know in your review you stated that using gamma 2.4 may be preferable on the AW. It does help with near dark tones, but without colorimeter correction the rest of the gamma curve is pretty off. I understand this is down to a difference in how each panel is calibrated form the factory, but I much prefer the look of the MSI out of the box.

In HDR, I unfortunately see similar behavior. I’m using the HDR peak 1000 mode on both panels (creator for the AW, and Pro-User on the MSI). The AW has a much higher perceived gamma of the image, which makes things just appears way too light overall. Dark detail is again significantly lifted, and the overall image appears less vibrant and more muted. Colors overall are way less intense in HDR vs the MSI. Night time scenes in games look way too bright, and indoor areas that should be dark are more dominated by grayish tones. Overall I found it to really hurt the HDR performance of the panel. Even though black is still true black, the near black tones are lifted to a degree where it almost reminds me of an LCD panel in spots. I’m not sure if this is an issue with my panel vs the one you tested, but I did observe the same near-black lift on the AW3423dwf, albeit to a much lesser degree.

HDR on the MSI is much better IMO. Perceived gamma seems to be what I’d expect from something tracking gamma 2.2. Near dark tones are appropriately dark, Night time scenes feel like night. Colors have lower gamma and feel more vibrant as a result. This isn’t without its faults though- there does appear to be some over saturation issues with the MSI. Certain colors, particularly red, feel a tad oversaturated. This is not the case on the AW, where saturation feels appropriate, its just that overall gamma is too low. On the MSI, gamma feels “correct”, but saturation seems exaggerated in spots. I sort of like this however, as it adds a bit of extra vibrancy to color, which was probably done for the target gaming audience who tends to complain about color being “washed out” in HDR (usually because they are used to over saturated SDR content, like when wide gamut is used). Its nowhere near the oversaturation you experience when using wide gamut in SDR, I don’t want to give that impression- its just a subtle bit of extra saturation that I think works well.

The main takeaway for me, and the one that really differentiates the panels in my experience, is that the MSI feels more contrasty, with darker near black tones and more “correct” gamma in HDR, while the AW feels overall just way too light, with too low gamma in dark tones. The MSI is much closer in HDR PQ to my LG C2 woled panel, if that gives a frame of reference.

Build: the AW does indeed feel better built. However, its also more bulky and contains an active cooling fan, where the MSI is super thin with a heatsink for passive cooling. Being 40 now, I have to say I prefer the look of the MSI as its more understated. It looks sleek and thin, where the AW looks fairly gamery.

Overall: I like the MSI better. To me, it comes down to this issue with perceived gamma. It negatively affects the picture quality of the AW for me. Again, the AW3423dwf also had this issue to a degree, but it seems worse on this panel. The only thing I wish was better on the MSI would be color saturation in HDR, which I found to be a tad oversaturated. Overall however, I think I will be returning the AW and keeping the MSI.

If you read all that, thanks! And again, big fan of your work for several years!

[PCM2]

Hi Lockjaw333,

Thanks for sharing your extensive feedback on the two monitors – that’s a great addition to this thread. Also glad to see you’re a long-time fan of the reviews. What this highlights for me is an ‘interesting’ and somewhat concerning variability in AW3225QF calibration. Assuming there’s not something else odd going on like it enforcing some sort of ICC profile for you under HDR (and assuming you’re not using HDMI 2.1 as with some GPUs that may change the HDR representation, though I didn’t observe this on my system). If I observed what you had observed I would definitely want to keep the MSI in favour of the Alienware! Elevated dark shades so detail is less masked than it should be (an artificial uplift of detail, if you like) is a pet hate of mine.

My unit is definitely not calibrated to strict ‘sRGB’ gamma using ‘Creator = sRGB’. I know some other samples are – TFTCentral’s was – but my unit showed good ‘2.2’ tracking for dark shades overall without more than a mild gamma reduction. The MSI was the one with the elevated detail levels using its sRGB mode on the unit I tested (per my post). The Lagom Black Levels test shows strong blending of the first few blocks on the Alienware, as you’d expect to see, and doesn’t show a clear uplift of detail for the first row as you’d see with ‘sRGB’ gamma. Under HDR the AW3225QF that TFTCentral tested also showed good dark shade PQ tracking in their review and also their more extensive and specific testing exploring luminance accuracy across a broader range of APL window sizes. The first table in that article shows the ‘Peak 1000 mode’ under HDR where the errors creep in for medium-bright shades but not dark shades. What you’re describing under HDR sounds a bit more like how the Dolby Vision (‘Bright’ or ‘Game’) setting was presented on my unit, for HDR10 content. So a few things to check, just in case:

– What GPU are you using? Is it Nvidia or AMD?

– How have you connected the monitor up?

– You definitely have ‘Dolby Vision’ disabled in the OSD?

– You aren’t using any ICC profiles under HDR (even ones that may have been activated without you realising)?

– You mentioned using ‘Creator’ under HDR. The presets shouldn’t have any effect and you should have a message telling you this if you try to change them under HDR. But just in case something is being messed up here, try selecting ‘Game 1’ before entering HDR and see if it changes anything.

[Lockjaw333]

Yeah, it does seem your experience with the gamma performance of both panels is the opposite of mine. I did take note of how you said the MSI had somewhat lifted dark shades- definitely not what I experienced!

I would hope its not unit variability causing such a degree of difference, but I guess its possible. Being as expensive as these panels are, its not really feasible for the average consumer to get more than one unit on hand to compare. I wonder if its worth trying to exchange my AW unit in store? I did purchase from a popular electronics store here in the US, so I could exchange easily enough, but again I’m not sure if its worth it.

To answer your questions:

-I’m using an Nvidia card, specifically an RTX 4080.

– I’m using HDMI 2.1 for both PC and PS5 connection. I will try with DP and report back.

– Yes, Dolby Vision is disabled. In fact, when its enabled, near dark gamma looks similar to when its disabled for me.

– No icc profiles active under SDR. For HDR, I did run the Windows HDR calibration tool, but I’ve tried it without that profile (which simply sets peak nits), and there is no difference in near black tones.

– Yes, Creator for SDR. And just in case that somehow affects HDR in HDR peak 1000 mode, I figured I’d mention it. I did try settings it to Standard while in SDR and activating HDR, and there was no difference- still lifted dark tones in HDR.

I’ll circle back after I’ve tried DP, though that won’t help for PS5 unfortunately. I doubt it will make a difference but lets see.

I would love for this to be due to some sort of setting issue. If that’s the case and I could fix the HDR gamma issues I would likely keep the AW…

[Lockjaw333]

So I tried with Displayport, and I observe the same behavior. There are no icc profiles affecting it, and using different SDR presets before enabling HDR does not change things.

Unfortunately it seems like a deficiency in the calibration of my unit. I have seen a few instances on reddit of users saying the same about poor near black gamma performance and lifted shadow detail. I’m not sure if its a change in production at some point, or just due to variability, but I certainly am not seeing the same near black performance that some other review sites have documented in HDR. For lack of a better term, the overall picture in dark scenes just has an overwhelming grayness to it due to the lifted lower shades.

My MSI on the other hand is borderline perfect. One indicator to me is when I switch on HDR at the desktop level, I do not notice much of a change at all in the display, which means that SDR levels are being displayed correctly under HDR for the most part. On the contrary on the AW, when I enable HDR everything gets brighter and lighter, and near black tones, like on the lagom test site, get even more lifted than they already are under SDR-creator mode sRGB gamma 2.2.

Sadly, I think I will end up returning the AW. I think overall its a sightly more polished product with better build quality, but in the end picture quality is what matters most to me.

[PCM2]

Argh that really is a shame, I was hoping there’d be a simple fix. But I appreciate you bringing this odd variability to my attention anyway, it’s certainly something people should be made aware of if they’re considering the Alienware. The only other thing I can think of (and it might not be this) is that there seem to be some issues with some models with HDR performance on the RTX 40 series that doesn’t exist on the RTX 30 series. I currently use an RTX 30 series GPU for testing purposes, so does Tim from Hardware Unboxed and so does Simon from TFTCentral. I’ve received a few complaints of an oddly washed out appearance to HDR (as if there’s some sort of ‘fog’ from perceived gamma being too low, perhaps) on some ASUS models such as the PG27AQDM – and they overwhelmingly come from RTX 40 series users. So I wonder if others who are having issues with HDR on the Alienware have an RTX 40 series GPU?

[Lockjaw333]

So I rolled the dice on an exchange, and I’m happy to report that this unit does not have the extremely lifted dark detail and abnormal low level gamma that the first unit did. In fact, it looks a lot like my previous aw3423dwf did.

HDR no longer has this gray haze to it due to over brightened dark tones. Things look correct for the most part. Very happy, though it is curious why the first unit was so far off!

Still compared to the MSI, near black gamma is lower/brighter. Though the color on the MSI is still oversatured in HDR, while the AW looks pretty perfect.

Given that this new unit has normal gamma and color behavior, with the welcome curve I think I do prefer it now. I also got a nice bonus in that the AW32 is was on sale over the weekend for $100 off the normal price, so I got that difference back on exchange.

Both are amazing monitors. The extra $250 for the AW32 vs the MSI is hard to justify, so if you can find the MSI, you’re not missing anything.

Again, happy camper. Looks like it was a one-off in unit variation, as you suspected.

[PCM2]

Awesome – really glad to see your new unit is better and it all ended up well for you in the end. 🙂

[Anonymous]

Hello,

I am seriously thinking about buying either of these two Monitors and have two questions.

I noticed in Rtings’ comparison of them, the following:

1) The AW3225QF has much worse VRR Flicker than the MSI.
2) The 321URX has much worse out-of-the-box Color Accuracy than the Dell.

Could you two guys (or anybody else that has owned either or both of these Monitors) give us your thoughts on these issues?
Are these problems noticeable?

Thanks,

[PCM2]

1) It doesn’t. RTINGS uses a synthetic test which triggers flickering with very specific refresh rate range jumps. The level of flickering exhibited during that test could vary depending on the calibration (and certainly gamma tracking) and you would find with different refresh rate ranges or fluctuations tested the opposite monitor could flicker more. I’ve been noticing and assessing VRR flicker on monitors since 2018 and the MSI and Alienware were very similar in their behaviour, as you’d expect.

2) Inter-unit variation plays a role here. Their Alienware unit was closer to 6500K and had significantly better gamma tracking ‘out of the box’ (with the sRGB setting) than the MSI, offering superior colour accuracy. The recorded ‘out of the box’ sRGB colour accuracy also showed my Alienware to be slightly superior to the MSI I tested, but to a less significant degree with an average DeltaE of 1.22 on the MPG 321URX and 1.02 on the AW3225QF.

[idiot123]

Hi,adam.I read your review about the AW32, which mentioned font rendering issues. I would like to know how the font rendering on MSI compares to the AW32, as this is crucial for my purchasing decision. Thank you!

[PCM2]

Hi,

There isn’t a written review for the AW3225QF, so you must’ve seen that somewhere else or mentioned in the video review? It’s covered in the same section of the MPG 321URX review. The (minor) fringing issues are related to the panel technology and aren’t specific to one monitor or another using the panel.

[Author]

Posts