The ASUS PA32UCDM for professional colour grading
For a very long time, I have been waiting for a display suitable for professional Colour Grading with an entry-level price point. As a professional Colourist and Display Calibrator, I work all the time with the established industry standard displays like the Sony BVM-HX310/3110 or the EIZO Prominence CG-3146 and the new CG1. For my home lab, however, I never considered buying any of these displays for myself, because of the high price. Back in 2022, ASUS already released the excellent PA32DC OLED, which has the potential to be calibrated really well, albeit for SDR only. For around 4000€ (2500€ for the 2023 released PA27DC model), it is a very good alternative for the trusted Sony PVM-A250 OLEDs as long as they are properly calibrated. In April 2025, ASUS released the PA32UCDM QD-OLED for just around 2000€, which is less than 1/10th of the price of a BVM-HX310. This made me finally pull the trigger and buy one for myself. If this display is in SDR anything like the PA32DC, and has some additional value in HDR, it would be a real bargain - especially compared to the FSI-XMP310 for around 10000€, which has a similar panel. ASUS did not sponsor me in any way. I paid for the display myself; all opinions in this article are my own.
There are already quite a few reviews about the PA32UCDM out there which go into great detail about all the features and connectivity. In my review, I will focus on the colour accuracy of the display, as this is usually the part I find lacking in most of the display reviews online. ASUS states 1000 nit peak brightness, 99% P3 coverage and a <1 DeltaE accuracy. Without any first-hand testing, I never trust any of these kinds of marketing claims.
Included in the Box is a factory calibration report which at least provides some reassurance of the display’s accuracy and potential. However, the report is based on very little measurement data. To prove the <1 DeltaE claim, they included 3 bar charts for P3, Adobe RGB and sRGB. Each chart shows the DeltaE values of 32 measured grey values and colours. It is not stated which exact RGB code values were used, but if the calibration reports from the ProArt Software use the same values for verification, only primary and secondary colours were measured. So the <1 DeltaE claim is based on a check profile consisting of just 8 grey values and 24 colour values of primary and secondary colours. That pretty much only tells us that the display can hit the edges of the colour spaces with a high enough precision and that the greyscale tracking is somewhat accurate and that a very small number of primary and secondary colours track according to the colour target, but it does not tell us anything about all the other colours in the bulk of the 3-dimensional colour volume. In colour grading, it is very important that all of the colours in the colour space track correctly according to the desired colour space specification. Displays can exhibit non-linearities in the gamut mapping, which a small check profile verification may not pick up. It is also possible to optimise the autocalibration in order to show very good results for just a few colours in the entire colour gamut in order to make the report look better. This is the reason I always use a verification patch set of at least 1000 colours to assess the colour accuracy of displays. On the factory report under “Test Equipment”, we see “Minota Color Analyzer CA410”. This is a typo - the CA410 is from Konica Minolta. This device is a tristimulus colourimeter, much like my CR-100. The report does not state whether this CA410 has a probe matching preset from a reference Spectroradiometer for that Display Panel type and if it has one, the report does not specify which reference Spectrometer was used to create the probe correlation preset.
Calibration out-of-the-box
So, naturally, the first thing I did after unboxing was select the factory default preset for Rec709 with a pure power law EOTF (in SDR from hereon now “Gamma”) of 2.4 and 100nits peak brightness and ran my standard calibration verification colour patch set consisting of 1090 colours distributed across the whole colour volume.
As we can see in the report above, the colour accuracy directly out of the box does not match what ASUS claims. I never really expected that the factory calibration would be bang on the money, but it was still a bit surprising how far off this particularly important preset was out of the box. The included ASUS calibration report states that the measurements for the unit were accurate at the time of manufacturing, which was January 2025. Since then, it undoubtedly underwent some heat cycles on its way in a shipping container from Japan to Europe, where it then also probably sat some time on a shelf and inside trucks and trains until I purchased it in April of 2025. New displays also tend to drift a bit more directly after manufacturing since all the internals need some time to settle. Before taking the verification profile, I correlated the measurements of my CR-100 to my JETI Specbos 2501 HiRes reference Spectroradiometer. Peak Brightness was at 95 nits, and I set the reporting tool to calculate all the DeltaE values relative to that slightly lower peak brightness - otherwise, the average DeltaE 2000 would have been even higher than 1.63. By using the 95 nit as reference in the first out-of-the-box verification, the report shows us more clearly that for some reason, the greyscale tracks until 80% signal level slightly darker than Gamma 2.4 and above 80% it tracks brighter than Gamma 2.4 with an average of Gamma 2.447. This deviation in EOTF tracking is odd for a factory-calibrated display. Colour Primaries and Whitepoint could also be more accurate. White balance was a bit on the yellow side; this deviation alone was responsible for a big part of the 1.63 average DeltaE 2000. The relatively stable drift graph in the bottom right corner is nice to see, the display stays relatively stable while cycling through all 1090 colours of the verification profile (at least in SDR).
Results after ProArt autocalibration
Next, I used the ASUS ProArt Calibration Software to recalibrate the Rec709 factory preset. I used ColourSpace and my 1090 colour verification set to verify and illustrate the results. The colour accuracy is already much better than before. Autocalibration takes care of the deviations in peak brightness, the white balance, the EOTF tracking, and the gamut mapping. The average DeltaE2000 came down to a respectable 0.6 with with 968 patches measuring DeltaE <1, 115 patches between DeltaE 1 and 2.3, and only 7 patches above 2.3, with one point reaching the highest deviation of DeltaE 2000 of 2.8. This is now in line with the marketing claims from ASUS of a <1 DeltaE accuracy - and this across the whole colour volume and not just for 32 colours, which is very good for an automatic calibration procedure! These results put the PA32UCDM roughly on par with a used (5000-8000h runtime) Sony PVM-A250 OLED Display. It has to be noted: although the ProArt autocalibration was used, it achieved these results with high-end calibration hardware that not everybody has access to. Also, the Konica Minolta CA410, which ASUS used for factory calibration, is a very expensive high-end colourimeter. The ASUS Pro Art colour space characterisation process measures a lot of very dark colour patches near absolute black, and the patches are very close to each other in the shadow part of the profiling. This is necessary to get the EOTF tracking as accurate as possible. I am not sure if more budget-oriented colourimeters would be capable of producing reliable enough readings to produce good auto-calibration results. I have an X-Rite I1 Display Pro still in a box somewhere, but I did not try it on the ASUS since this colourimeter is already 10 years old, and I would not trust the readings of such low-cost devices after they are older than 2 years. But I am curious to know what results a newer budget-oriented meter can achieve with the ProArt autocalibration feature. However, the full potential and accuracy of any Display can only be unlocked with high-end spectrometers and colourimeters (and some experience). It is also the only way to match different display types in the same room/facility.
Impressions
When calibrating the white point with the reference spectrometer, it was also nice to see that the greyscale perceptually matched really well to my 13” iPad Pro (M4, 2024). And everybody who read my last article knows that I keep my iPad Pro perceptually matched to a calibrated BVM-HX310 whenever I get the chance to do so. I use that as a form of visual sanity check. The match can be improved with custom D65 white point colour coordinates specifically for the ASUS QD-OLED Panel, but for this kind of perceptual fine-tuning, it is necessary to create custom presets which incorporate the perceptual match directly within the calibration LUT. The colour coordinates of the factory presets can not be changed; any perceptual fine tuning there would have to be done through adjusting the R/G/B Gain sliders after the autocalibration, which I do not recommend. Please note: for the photo above, my phone camera’s sensor and its automatic image processing, the difference in white balance was stronger than for the human eye.
After looking at a few hundred test images, real life and synthetic, the colours on the ASUS stay pretty stable when viewing the display off-angle. There is also hardly any decrease in brightness and contrast when viewing the display off-angle. It is surprisingly good. In comparison, there is a considerable drop off in brightness when viewing an iPad Pro 13” (2024) off-angle. The off-angle colour shifts on the ASUS are less pronounced than what I am used to from Sony PVM-A250 OLEDs or even the Sony BVM-X300 OLED. The off-angle viewing colour accuracy of the ASUS is probably not quite as good as the latest BVM-HX3110, but far superior to the BVM-HX310. The ASUS PA32UCDM has a QD-OLED Panel similar to the FSI-XMP310. I calibrated a few XMP310 already, and people who have experience with it know that in dark images sometimes artificial vertical or diagonal stripe patterns can occur. In some content, these are visible even from the normal viewing distance and can feel quite distracting. One of my concerns when ordering the PA32UCDM was that this display would have the same issue. I tried to provoke these patterns with the same test images that brought them out on the XMP310. I had to go much closer to the ASUS display to be able to see the Artefacts. In the normal viewing distance (for a 32” Display, that would be between 60cm and 120cm), I found them to be a non-issue for me on the ASUS, which was a positive surprise. Also, the ASUS does not add any artificial noise to dark image areas, and dark grey patches look clean. Gradation is also smooth, there are no visible steps on grey ramps and colour ramps, which indicates high-quality LUT interpolation on ASUS PA32UCDM. Also great to see: there was no distracting gamut reduction for very low signal levels on the ASUS. This was another serious concern I had before purchasing the PA32UCDM QD-OLED. The FSI-XMP310’s Gaia autocalibration visibly reduces the gamut of low signal levels - dark colours would almost look grey compared to what they look like on reference displays like the BVM-HX310. The ASUS PA32UCDM does not have that issue. The ASUS QD-OLED Panel has a glossy surface with an anti-reflective coating. In colour grading suites, the feeling you get from the display will be very close to any other high-end grading display with true black capability and a glossy surface. If there is some light in the room, the anti-reflective coating will slightly reduce the contrast, which manifests in perceptually raising the black a little bit. This effect can also be observed in some QD-OLED TVs on the market. Screen uniformity is also good on my unit, without any distracting colour shifts towards the edges of the screen.
Results after Profiling and Custom calibration LUT upload
One major selling point of the PA32UCDM for me is, of course, that it is fully integrated with Lightspace Colourspace. There are 3 custom preset slots available. By bypassing the Pro Art Calibration Routines, profiling the display with ColourSpace and uploading custom LUTs into the display, it is possible to improve the accuracy of the device even further.
Now, the average DeltaE 2000 is down to an impressive 0.33 - a substantial improvement over the already good ProArt autocalibration. With this approach, there are no colours above DeltaE 2000 of 2 anymore, with the highest deviation being 1.6. 97.8% of the measured colours were below DeltaE 2000 of 1. This was only the first try, and I see still room for some improvement. After optimising and adjusting my calibration workflow for the ASUS QD-OLED Panel, it should be possible to get even slightly better results than that. But even now, the ASUS PA32UCDM for Rec709 Gamma 2.4 SDR is on the same level as the EIZO Prominence CG-3145/46 in terms of colour accuracy. For a device that costs only around 2000€, the ASUS performs outstandingly in SDR. Only the Sony BVM-HX310 is even better ( <0.25 DeltaE 2000 typical for Sony BVM-HX310).
HDR
When it comes to HDR, we get to see some shortcomings with the ASUS PA32UCDM. ASUS claims 1000 nits peak brightness, but this can only be achieved on a very small display area (3% window). As a result of that, ABL is triggered quite easily. And that is a bit of a shame since the QD-OLED does not have the issue of gamut contraction for very bright colours, like the W-OLED Technology has. In the factory presets for HDR, it is possible to influence the ABL behaviour. There are 3 modes called “basic”, “clip” and “optimised”. Basic and optimised are 2 different ways of highlight rolloff, whereas clip just clips the brightness of all the pixels which the display processing deems as too bright to display accurately, whereas the rest of the image remains accurate to the colour target. In colour grading, “clip” is the preferred mode, but with the ASUS PA32UCDM, the ABL will probably be triggered a bit more often than users would want to. As soon as you see clipped highlights, you are overpowering the capabilities of the display. There is no warning light when ABL is kicking in, like, for example, on the BVM-X300 OLED from Sony. The ASUS Pro Art autocalibration does a decent job of calibrating the HDR presets. It is also possible to improve the calibration by profiling the display in HDR and create a calibration LUT (which is quite a bit trickier to get right, compared to SDR). However, the 3 ABL modes are not available on a custom HDR preset. So the colour accuracy of a custom mode can be better than a respective autocalibrated factory preset in terms of what the calibration report shows in numbers. But if in HDR custom mode the ABL is triggered, the brightness of the whole image is reduced, and not just the pixels that are too bright to display. Another thing to consider is the relatively low thermal mass the display has at its disposal to dissipate heat quick enough. For consumer devices it is of course desirable to have a thin and lightweight design. But in HDR that leads to the problem that heat from very bright image areas can not be dissipated quickly enough. This makes the display less thermally stable, and colours will drift slightly depending on the screen temperature.
The Graphs above show some colour drift going on while cycling through the 1090 colour verification patch set in HDR brightness. On a High-End Dual LCD type colour grading display with lots of thermal mass and active heat management, there would be almost no drift during a verification run as long as the device had enough time to reach steady state after switching into HDR. This is why I recommend the ASUS PA32UCDM primarily for colour-critical work in SDR (where heat cycling is much less of an issue) and only for occasional, non-critical HDR work. For absolute colour-critical work in HDR on an everyday basis, the high-end dual LCD colour grading displays from Sony and EIZO are still the better choice.
other (small) weaknesses
Now let’s talk about the things I do not like about the ASUS PA32UCDM. The following issues were observed with Firmware MCM103 and MCM104. What bothers me the most is ASUS's use of terminology. For example, the “HDR_PQ DCI” preset: just from the label, it is unclear what this setting means other than that a PQ EOTF is used. Only in the ASUS ProArt Calibration Software is it stated that for this preset, P3 Primaries and a whitepoint of “6500K” are used. However, when measuring the whitepoint, it is D65, which is luckily the whitepoint that I want for this preset. I do not know why ASUS and some other manufacturers (EIZO, FSI) equate “6500K” to D65 or “6000K” to D60 and so on. These two units are not interchangeable and have different meanings and definitions. The colour preset mentioned above should be named “HDR-P3D65-PQ” instead. This would leave no room for confusion. In the factory preset for “DCI-P3”, it is the same issue. The name suggests a P3 colour space with DCI whitepoint, but a submenu in “Pallette” gives 2 more options for the whitepoint, the default being “6500K” and the other is the actual yellowish/greenish DCI Whitepoint (x 0.314, y 0,351), which they call “P3-Theatre”. Why not call the preset “SDR-P3-Cinema” or “SDR-P3-48nit” instead, with the whitepoint submenu offering “D65” and “DCI” as options for the whitepoint? The respective peak brightness for each factory preset is also only stated in the ASUS ProArt calibration software. It would be nice if they were right in the label in the onscreen display. The German translation of the ASUS Software and the onscreen menu is also not very good and even further increases confusion. It is better to set up the display in English instead. I can’t say anything about the other language options. The Setting for the input signal range is, unfortunately, a global setting and is applied to the HDMI and the USB-C inputs - it would be better if each of the inputs had separate settings for input range. When switching between a factory SDR and HDR preset and vice versa, one has to make sure that in the SDR presets “Uniform Brightness” is always ON and in the HDR presets “uniform brightness” should be OFF if brightness levels above 250 nit are required. Remembering to set this correctly whenever switching between HDR and SDR is crucial! If “Uniform Brightness” is OFF in SDR factory presets, the peak brightness varies depending on how many bright pixels are displayed. The “Uniform Brightness” option should be individually adjustable for each preset and not a global setting. Custom SDR user modes, which have been calibrated via a calibration LUT, seem to set the “Uniform Brightness” option automatically to ON and grey out the switch, which makes the user presets less prone to operator errors. There also seems to be a bug: When adjusting the “black level” in a preset the numerical value seems to be saved according to the onset menu but when switching to another preset and then back, the black level has to be reconfirmed to the desired value - it is not properly loaded on preset selection. Occasionally, I found that the Display does not properly recognise a 10-bit per channel input signal and would truncate to 8-bit. This goes mostly unnoticed in SDR but would be picked up visually in HDR quite easily. Switching the Display OFF and ON again is one workaround for that issue. Another small issue, which only impacts people like me as a calibrator, is the position of the onscreen menu and the position of the control buttons. The Onscreen menu is at the bottom centre of the display, but reaches right into the very centre of the screen, where ideally the probe would take its readings. So every time the menu comes up, the readings are going to be wrong. This is especially annoying when I try to manually set the white balance and peak brightness of the screen. While navigating through the menu and changing the R/G/B Gain values, I can not take probe readings from the centre of the screen. The control buttons in the centre of the bezel are also exactly where the tripod with the probe is standing, which makes it very inconvenient to operate the display menu while taking measurements. When connecting the Display via HDMI to a BMD Ultrastudio 4K Mini, it seems to be best to always send the full 16:9 Aspect Ratio to the ASUS Display (for example, 1920x1080 or 3840x2160). If the Video output Pixelraster is set to scope Aspect Ratio (for example 3840x1600, 2.4:1), then the image does not get centred on the ASUS screen, instead, it gets pushed to the bottom of the screen with just one, twice as large letterbox on the top of the screen. This is a little annoying since scope Aspect Ratio content with a 16:9 output Raster will always show black letterboxes around the content in the Resolve viewer - and I prefer the viewer only to show the active image area.
Although some of the listed issues are a bit annoying, none of them are dealbreakers for me. The display has it where it counts: An excellent panel that, with proper calibration, rivals even the most expensive reference colour grading displays in SDR. Only the very strong ABL in HDR prevents it from being a drop-in replacement.
And hey, it comes with a free stand - nothing to complain here.
Party Trick
The USB-C port of the iPad Pro can be used to connect an external display. The image above shows the iPad Pro receiving a video stream via DaVinci Remote Monitor from my Resolve Laptop over WiFi through a VPN tunnel and displaying the video stream on the ASUS PA32UCDM. I will need some extra time to figure out the correct setup between iPad and ASUS. Since Apple iPad OS and Colour Sync are involved, I will have to separately verify the colour accuracy of this use case and update this article as soon as I know more.
Verdict
The ASUS-PA32UCDM is an excellent entry-level colour grading display for SDR and rivals the most expensive reference Displays in terms of colour accuracy, provided it is properly calibrated before use. I wish it were better in HDR, but when high Brightness is required, it can only provide that on a very small screen area. Since a thin and lightweight device is preferable for most consumers, ASUS had to make some compromises in HDR. A heavier design with more thermal mass and better active heat management would have resulted in much more usability in HDR (less ABL, more screen area with max brightness) and even higher accuracy in SDR due to less temperature-related colour drift. The Display and firmware still have some quirks and weaknesses. I hope some of these will be addressed in future firmware updates, but these small issues can not take away from the excellent colour accuracy this device is capable of delivering. Colourists who just look for a good display, are ok with just one HDMI connection and do not need any of the professional connectivity options and additional features that more expensive displays offer, should take a closer look at the ASUS PA32UCDM.
