Has the Alienware AW3423DWF Improved? - Firmware Update Tested

welcome back to monitor's unbox today we're checking out the latest firmware update for the Alienware aw3423 DWF to see if it's fixed the inaccurate hdr-1000 mode that kept this monitor from being an overwhelming recommendation this update firmware version m3b103 has been available for a few days now from Dell's website and it's pretty easy to install if you follow the included instructions but first a quick recap of why this monitor needed a firmer update select the other QD OLED monitors that I've reviewed so far the aw3423dwf ships with multiple HDR modes and of course HDR is very important on this product as it's the main reason you'd forecast over one thousand dollars US for it the two modes that are of primary consideration are the display HDR true black mode and the HDR Peak 1000 mode which is supposed to differ in their brightness capabilities the true black mode should top out around 400 nits while the peak 1000 mode show would unlock the ability to push up to 1 000 it's a peak brightness for some elements what I found in my review is that the true black mode is reasonably accurate not perfectly accurate but it does an okay job of following the eotf curve meaning that whenever HDR content tells the monitor to Output a certain level of brightness the monitor in this mode delivers roughly that level of requested brightness however as expected you're not going to get much more than about 450 nits out of this mode the peak 1000 mode was a different story while this mode does push brightness up to around 1000 it's for small elements around a two percent window size for example the entire eotf curve is inaccurate actual measured performance for a given brightness is too high meaning that content is brighter than it should be for pretty much the entire brightness range this is quite a noticeable difference in fact in several circumstances this mode delivers double the expected brightness of the content at hand which prevents the monitor from delivering as Rich Shadow detail as it can Nvidia GPU owners also had a few issues with the monitor producing washed out colors unless the display was switched into the console mode with Source tone mapping set to on what Dell needed to fix was the HDR Peak 1000 modes eotf curve so that it delivered accurate brightness throughout the entire range while still preserving the ability to hit 1000 nits of brightness this would give it performance parity with the aw3423dw the g-sync version of this monitor as well as the ray recently released MSI 342c unfortunately though firmware version M3 B103 is only a partial fix for this problem and in fact has a few performance Oddities that require a more detailed examination testing the peak 1000 mode in the exact same way that I tested the original firmware of this monitor revealed some performance improvements but this was restricted to the lower end of the luminance range for the rest of the range indeed the majority of the eotf curve the DWF still exhibits raised brightness to the point where elements can still be doubled the expected brightness pretty obvious in games if you just flick between the true black and Peak 1000 modes you'll see the peak 1000 mode brightens the entire image which is not correct Behavior what should happen is that the image remains largely the same but highlights extend to 1000 nits and detail for higher brightness levels is improved this is exactly what happens on the more accurate DW model which doesn't have this issue so basically this firmware update does not fundamentally fix this issue and Alienware still needs to put in some work to deliver an accurate Peak 1000 setting however while this isn't something I tested extensively in my initial review I did notice that the differences on an Nvidia GPU in the console mode between Source tone mapping on and off has reduced there is still a difference it's just that the default configuration is now a lot less likely to be washed out than the original firmware which is a good thing while I was able to confirm the peak 1000 mode is not fixed in this firmware update the true black modes accuracy has improved and now looks very solid those improvements to the lower part of the eotf curve have tidied up performance and this mode Remains the best configuration for this monitor just with you know disappointing Peak brightness with this in mind and discovering the firmware isn't ideal I embarked on an investigation to see whether the dwf's peak 1000 mode can be improved through other means I've seen on various places including Reddit that people have come up with several supposed solutions to the problem so I decided to put them to the test which brought me down a bit of a rabbit hole of testing and ended up spending far too much time trying to crack the puzzle but here's where I've gotten to so far the crucial finding I've made is that the way this monitor behaves is definitely different depending on whether you are using an Nvidia or AMD GPU the general performance characteristics are the same in its default configuration especially with Source tone mapping on for NVIDIA gpus so to be clear it's not the case where you get accurate performance on an AMD GPU but inaccurate performance on an Nvidia GPU they're both pretty similar but the way some monitor settings behave for NVIDIA owners is different to AMD owners and this impacts the final performance users can achieve the end result is that if you own an Nvidia GPU it is possible to improve the peak 1000 mode into a largely usable State via a few adjustments but if you own an AMD GPU these tweaks are ineffective and it's not possible to fix the peak 1000 mode quite an interesting end result that I rarely see with monitors almost always performance is identical whether you're using Nvidia or AMD but in this case in the HDR mode that's not true and I believe it has to do with the tone mapping implementation on AMD cards it uses the freesync 2 pipeline which is now called freesync premium Pro whereas on Nvidia gpus it uses a normal HDR pipeline without spending ages in this video explaining the differences the basics are that these two implement differ in where and how they turn map HDR content and how much information is passed between the display and the GPU on this particular monitor there seems to be some inconsistencies between how these pipelines are implemented the causes of performance and setting difference depending on whether freesync premium Pro is active or not which in this case is whether you're using AMD or Nvidia and as far as I'm aware there is no way to disable the HDR portion of freesync on AMD GPU you can toggle whether freesync's adaptive sync implementation is on or off but not the tone mapping stuff though if someone has an idea on how to disable it I would love to hear it so on Nvidia gpus the way to get a more accurate 1000 experience involves three simple adjustments turn the console mode on turn Source tone mapping on and reduce the contrast to around 64. with these setting changes implemented I was able to correct eotf tracking to a more reasonable degree significantly reducing the amount of over brightening it's not perfect though you can see that brightness roll-off isn't great in the upper range but it definitely works better under this configuration and when we test the monitor using a two percent window size instead of ten percent we can confirm that tracking is still better but we retain Peak brightness at under a thousand nits which is also what we saw initially and what you'd want from a 1000 nit mode the main issue with this sort of configuration is for high window sizes when we get up to say 25 there are issues with brightness roll up above about 100 nits surprisingly we saw a similar issue with the g-sync model the DW model in its 1000net mode I still think this performance is better than the default over brightened experience but it's not ideal and a proper firmware fix would hopefully tackle this problem on AMD gpus this fix is ineffective firstly Source tone mapping is grayed out as it's using the freesync premium Pro pipeline instead and secondly reducing the contrast also reduces Peak brightness unlike on Nvidia cards if we reduce contrast to around 57 as an example we do see more accurate eotf tracking with reduced over brightening but this also reduces Peak brightness when testing using a two percent window from around 970 nits to just 530 nits which defeats the purpose of the fix if to get an accurate HDR experience we have to reduce brightness to just 500 nits we may as well use the true black HDR mode instead which is more accurate without needing to mess with other settings so basically AMD owners will need to wait for a proper firmware fix to correct the peak 1000 mode one solution that I heard from multiple people on Reddit was to adjust the display's edid HDR metadata values using Cru however after testing this on both AMD and Nvidia systems this doesn't actually do anything apart from changing their numbers some apps report I saw this listed as a common fix to increase the cutoff in the windows HDR calibration app from around 500 nits to around a thousand nits but this isn't really what's happening while changing the edit will change the number in the calibration app where this cutoff occurs the actual brightness performance is the same and whether you do or don't change the edid the monitor still sees a cutoff around 500 nits it's just that with the adjusted edit the app reports a thousand nits when it's actually still 500 nits and further confirming this using calman shows no actual performance difference and no change to eotf tracking so I wouldn't bother messing around with CIU here after all this testing and findings there are still question marks over why there is such a large discrepancy in UTF performance between the 400 nit and 1000 nit modes and why OSD settings impact the freesync and non-freesync HJ implementations in different ways the latest firmware doesn't solve the problems with the 1000 it modes to any significant extent though some further adjustments are possible for NVIDIA GP owners but I'm not a display engineer so I simply can't answer why we're seeing what we are seeing and how to properly fix it one thing that does appear to be happening is that the 1000 mode is using a rather crude scale factor for its eotf curve if I take the curve as reported in Cal man put it into Photoshop and lay it over the 400 nit mode curve then simply resize the graph down in height you can see the actual performance of the 1000it mode ends up matching the expected performance of the 400 nit mode it seems this mode is just stretching the entire eotf curve to achieve 1000 nits rather than extending the curve up to 1000 nits as it should and when we compare the DWF to how the DW and MSI 342c work in their respective 1000it modes there's a further difference in operation with the DW and 342c when you switch from the 400 nit mode to the 1000 nit mode the display's metadata changes to reflect the difference in capabilities and the display briefly resets it flicks to Black for a moment indicating that its processing is switching over to an entirely different processing method method with the DWF changing to the 1000 mode is instant there's no change to metadata no obvious switching into a different processing mode and as seen previously there's no proper adjustment to the eotf curve I don't know how these monitors work for sure again I'm not a display engineer and I didn't design them but perhaps the solution here is to have the DWF properly switch over into a different 1000it mode complete with a tuned eotf curve for 1000 in performance and updated metadata this of course would require another firmware fix and may require significant changes to how the monitor operates but judging by how the other QD OLED products operate it should be possible to make the peak 1000 mode more accurate in any case for now this firmware update doesn't significantly change my monitor recommendations and what I discussed in my initial review video hopefully those with Nvidia gpus will benefit from the tweaks I talked about and I do think it's worth updating to the new firmware but Dell still has some work to do before this product delivers an optimal experience anyway that's it for this video just a brief update on the aw3423dwf if we see further firmware updates that do fix this problem give us a better Peak 1000 mode we will dive into that a bit more but I just wanted to take a look at this firm update and also provide some of those you know tweaks for people with Nvidia gpus which I do think is something I didn't really cover in the initial review that I did and does benefit the experience for those with that product so yeah that's it for this one if you do appreciate our independent testing and investigations into these sorts of things even if it's not something that is going to be relevant for everyone this really is just for owners of that monitor then please do consider supporting us via patreon or flow plane links are in the description below anyway that's it for this one and I'll catch you in the next one [Music] foreign foreign foreign

Channel: Monitors Unboxed

Views: 98,712

Keywords:

Id: vzY7q1bSrWM

Length: 12min 57sec (777 seconds)

Published: Wed Mar 15 2023