The actual initial topic was the increasingly frequent question from colour grading colleagues, but also from production companies or post houses: How well suited are Apple’s latest iPads for colour grading acceptances that have the new “reference mode”? Reason enough to deal with both topics at once, as they inevitably belong together if you want to be able to answer the question of suitability. I looked at two current Apple devices: the iPad Gen 6 and the latest MacBook Pro – both with the Liquid Retina XDR.

Externals

What makes the reality even more complex is the question of whether to use the internal/own displays or other computer monitors. You could avoid the problems as far as possible if you used reference monitors on a video IO for colour grading control and not the internal graphics card output on internal or external displays.

However, using external monitors requires even more expertise: I was just called to a colour grading suite at ZDF where a colour grader was sitting in front of one of the most expensive reference monitors you can currently buy: a Sony BVM HX310. It was supposedly not calibrated. After all, the colourist noticed that the colours could not be correct (too saturated compared to the GUI monitor).

However, it turned out that the HX310 was not set correctly: instead of Rec709, the HX310 was on the P3 colour space. The Windows computer and Resolve with its GUI monitor were set to Rec709 output. Such a combination is normally less prone to errors compared to the Apple workflow, provided the users know what they are doing.

So what are the boring basics?

The studio standard, which is still the worldwide standard in colour grading and VFX workflows, is still Rec709 (identical to sRGB in terms of colours), only in cinema or HDR is P3 (the latter embedded and limited in BT2020/BT2100) common in delivery as a colour space, but by no means always used to its full potential.

It is not uncommon for Rec709 to simply be addressed and displayed in the P3 colour space or even embedded in BT2020. This can work well if you transform cleanly between these two colour spaces. This usually works in a colour-managed workflow, but the more settings that can be set, the higher the error rate. But this is the crux of the matter: users and colour management must both work well together ;-), especially if the monitor is set to P3, which has been the case with Apple’s Retina displays for many years and not just since the XDR generation.

Rec709

Let’s get back to the Rec709: The following decisive secondary conditions must be observed, which are often overlooked: The dear monitor gamma, which was unfortunately forgotten to be defined in the standard, but which was also more or less fixed in the days of CRT monitors: gamma 2.4. Only with cinema or today’s modern monitors has the gamma (become) variably adjustable to suit: in dark cinema 2.6 and in office light conditions (as preset on most computer monitors): Gamma 2.2.

It is important to note that the gamma is only and exclusively set on the monitor to match the ambient light and is never changed or adjusted during export. This is where the first application errors occur. Because theoretically an adjustment is possible and is therefore all too often done due to a lack of expertise and voilà: The export looks different for this reason alone.

If this is not noticeable, it is due to incorrect monitor settings and/or incorrect ambient light conditions. The latter are defined for Studio Gamma 2.4 = 10 per cent of the peak white level, which is 100 nits (nits = cd/m²). This is achieved by completely banning daylight from the room and using some D65 (6500 Kelvin) compliant white light (bias light), i.e. not using the usual warm light sources.

And now iPad?

If I want to use an iPad as a reference device, I have to create studio lighting conditions, because it is set to gamma 2.4 in reference mode. Using it in an office/daylight environment is therefore pointless. I’d rather use a better TV or graphics monitor that can be set to gamma 2.2, but also at least provides an sRGB or Rec709 preset.

At least you can’t set anything wrong on the iPad in reference mode. This could be categorised as Apple-typical “user-friendly”. It is more complicated on MacBooks, there is no “reference mode”, I can and should set and adjust the colour spaces correctly via the monitor settings as on a reference monitor, and at best also calibrate them.

The 6th generation iPad Pro is absolutely “reference-capable” when calibrated – because you can at least calibrate the white point on the iPad, as with any “professional” reference monitor. If the display has been well constructed and well calibrated, no more is necessary. However, this process is complicated and not really interactive. You need a calibration probe with an x,y output or display, ideally something like a Jeti 1511 spectroradiometer, which provides exactly that.

It’s a bit strange that you can’t enter 100 nits for the brightness, the setup only accepts values below that – but this is hard to tell visually if you only get 93 nits instead of 100. It took me several attempts to get the calibration to a good value. The software does not deliver the value that you set and that was measured. If you measure again after the accepted setting, you realise that the final value is slightly off. You then have to “shoot” just right to achieve a precision landing. After a few times back and forth, it worked. In my case, a reset to default improved things even more.

The background

So far it looks good, but here’s the thing: For a P3 display to show Rec709, a functioning colour management system is required. This presupposes that both the OS and the respective software are capable of communicating with the colour management. DaVinci Resolve and Apple’s own Quicktime Player are capable of this. However, for the Quicktime Player, certain flags (metadata) in the video files are also responsible for whether or not certain standards are used. And here Apple has (historically caused) a system design deficiency in colour management: they use an outdated gamma. To “correct” this, a special flag is required in the .mov/.mp4 file if it is to be played back correctly on Apple devices: NCLC does not have a transfer function specified in meta tags for Gamma 2.4 – so tagging “1-2-1” tagged for Rec709 Gamma 2.4 reads the “2” in the 1-2-1 tag as “unspecified”, so Apple Colorsync (the operating system) ignores it and applies the internal incorrect gamma.

The result

And (almost) every player on a Mac displays the file with gamma 1.96, which is a far cry from gamma 2.4. So people check their files on a Mac (which by default uses Quicktime, a colour managed player) and the files are displayed washed out with too flat gamma. However, this can be torpedoed and obscured if Fullrange vs Legalrange is used, which depends on the colour management/codecs and export setting. However, the standard is legal range (video level measured externally = 0-100 per cent – but be careful, the display of the internal waveforms in Resolve or other programs show what is processed internally, which is usually 0-100 per cent full range and not legal range, which is what is to be exported. So Rec709 is actually exactly Rec709, gamma 2.4, for 100 nits at 10 nits ambient light in D65 and white point D65 at legal range (note: legal range is at 8 bit = 16-235 or 10 bit = 64-940) When exporting from Resolve, the Rec709A setting can now be used instead of Rec709 – Gamma 2.4.

The NCLC tag

And now this leads us to NCLC tag 1-1-1. This flag does not lead to any burn-in of other gamma, but only to a correct display of gamma 2.4 on Apple OS operated colour-managed applications, if this has been set on the display – which is the case by activating the reference mode on the iPad Pro. A preset must be created for this on the MacBook. Minimum settings in DaVinci Resolve: Only amateurs use CSTs in Resolve, the professional who works as efficiently as possible uses DaVinci’s YRGB Colourmanaged CM. This allows you to define the target for the subsequent export and correctly address the metadata for the send files, especially if different colour space targets are required. This often occurs with Rec709&HDR in combination with today’s streaming platforms… So first set the white point and colour space optimally and then calibrate the white point. If everything is set and calibrated correctly, an optimised display preset must also be set in the OS. Here it is still set to P3-1600 nits: reduced to 45 per cent brightness:

Conclusion

This may come as a surprise, but Apple’s current colour management actually works – if it weren’t for the Quicktime flag problem, which Resolve solves with the Rec709A (the A stands for Apple) export flag (1-1-1). As this flag can ensure correct playback on Mac OS and has no effect on Windows, you should always export with it even in Resolve on Windows, then the files will also look correct on a Mac, even if the Mac laptop display is “incorrectly” set to P3 1600 nits. But you can achieve real perfection with a Rec709 display preset. Both the Macbook and the iPad Pro with the XDR displays are absolutely suitable for reference if the listed conditions are observed. However, if you use an external reference monitor with a Windows system or Mac system that is incorrectly set to P3, then the user loses out and an expert is unnecessarily called in to calibrate. In this respect, self-contained systems such as the current Apple devices are one level less error-prone. But as the social media forums impressively demonstrate: This doesn’t really help. “KnowHow is king” and that is here and not on social media

This will be a short review – not least because I’m really enthusiastic about the tool, which is admittedly rare.

Do you really need it?

After the initial praise, I have to say that I doubt whether the Checkr will be used as much as it deserves. I have referred to the use of colour charts in many workshops – but I have never seen it used in any production myself, at least none that I have been involved in. Especially if you colour correct yourself and don’t really need it – except when the tools in Resolve don’t always work as far as Resolve’s automatic recognition procedure is concerned. If you wanted to counteract this – at least until now – you had to use colour charts manually, but this was only time-saving to a limited extent and usually quite inconvenient. Unhandy, sensitive and not really standardised. This has changed with the new Checkr video. And before anyone says anything: that’s not a spelling mistake “Checkr Video” is the name Datacolor uses.

There are now also so many different colour checker boards that it can be tedious for the colour grader to select the right module in Resolve using the auto-recognition function. But with the Spyder colour checker, this can be done manually using waveforms and vectorscopes and with any software that has halfway usable grading tools, including Media Composer, Premiere, Final Cut, etc.

Use in front of the camera

Enough adulation, how do you use it? Ideally, we hold it in the main light source or next to the face; depending on the image section, it can also be held under the face (pro tip: be careful, the head casts shadows). It is important to say this loudly (and repeatedly) to anyone who touches a colour chart: Do not touch the colour fields themselves with your fingers, as the slightly acidic skin grease destroys the colours.
Once you have the colour chart in the picture, slowly wiggle it vertically and horizontally. This not only looks funny, but above all ensures that the shiny black surface is at least once in the picture without reflection, thus providing maximum black for contrast. If you’re shooting with studio lighting: put it on a tripod. And if you have to employ an extra useless assistant (someone’s nephew)..

Ideally, we should also not overexpose (or temporarily reduce the aperture when exposing high key). It is also important to mark the shot as a separate clip and make sure that these clips have been exposed once for each lighting situation for all cameras. You can give this to your colourist as a separate collection and save a lot of time when colour-matching very different cameras.

Use in post

The colour warper in Resolve is still quite new and is best suited for colour matching. In all other programmes that do not have one, the Hue vs. Hue and also the Hue vs. Sat controls should be used. However, the latter require more time (especially in Final Cut or Premiere) than the same tools in Resolve. The colour warper is also available as a plugin from Nobe – is.gd/color_remap – it’s really worth the money and we’ve been fans of it for years. If you want to know exactly what’s in the latest version, you can take a look at Uli Plank’s text from page 70 onwards. And to be fair: The 3D LUT Creator by Oleg Sharonov was the first to bring these functions “Conveniently” to the people – See DP 19:03, or in full text here: is.gd/dp_3dlc

Colour grading

Before we do anything with the frames, it is necessary to optimise lift, gamma and gain according to waveform. Colorwarper is then the tool of choice – in comparison, with Hue vs. Hue and Hue vs. Sat you have to use two different tools one after the other and alternately, where you also have to set the colour vector points first, which is particularly time-consuming in Final Cut.
In the colour warper, you can work directly and can already touch the vectors. Only if too many vectors were previously preset, then you should reduce them to six.

Conclusion

The Checkr belongs around the neck of every cameraman or AC or VFX supervisor. And as soon as several cameras of different types are involved, it generally does no harm even with VFX shots if no others are used!
But one more tip: If the part is regularly used outdoors, then you should get a new one every year or every two years (the Checkr costs around 150 euros), because these colours also fade over time, especially due to UV radiation in the sun!

The new Asus ProArt PA279CRV is a relatively affordable monitor (€550) that aims to give the impression of a premium professional panel. If you ask Co-Pilot of Windows 11, you get the following info: “The Asus ProArt PA279CRV is a monitor that is making waves in the tech industry due to its impressive features and affordable price. This 27-inch 4K monitor is part of ASUS’ ProArt range, designed for video editing and content creation.” It’s not just adverts that like to exaggerate, AI language models also get exuberant from time to time: “…making waves in the tech industry”

What interests us now about the Asus ProArt Display PA279CRV and also the Asus ProArt Display PA169CDV is precisely the topic: How much “pro” do you get for the money?

There are two stand variants on the back, one for a “Wacom-like” tilt, as a pen tablet, and one for a “high tilt”, as known from external monitors. In the high stand version, you can also work comfortably in portrait mode – or simply read the DP-Epaper in the correct format. I’m just saying.

What are these classes?

Our definition comes from colour grading. Here, anything that achieves a so-called class 1 (as far as possible) is considered professional. For example, a Sony BVM HX3110 (dual LCD) or an EIZO CG3146 or a Sony BVM X300 (OLED). There would certainly be more devices, but these are the most common and cost between 20,000 and 30,000 euros. Mid-range for me so far has tended to be an EIZO CG series. They are around 2000-4000 euros. Classic class two broadcast monitors are similarly priced or around double or triple the price – depending on the manufacturer, dealer and age. Anything under 1,000 euros would therefore not be mid-range. A good 4K/27-inch office monitor from Eizo already costs 600-1,500 euros. In this respect, the Asus is almost in the low-end price range. For the sake of completeness: A favourable
Office monitor in this size and resolution is around 200-300 euros, i.e. about half the price.

The claim

Anyone who edits images (as a rule) also wants to see colours that are as correct as possible. In recent years in particular, the problem has arisen that different colours have to be constantly discussed. Not infrequently with the customer complaining that the colours are too bright or usually “too red”. The cause is often the same: many monitors can now display so-called wide gamuts and are preset at the factory with these natively very saturated colours. If the operating systems or programs do not use colour management, the colours are often seen incorrectly because they were usually produced in sRGB/Rec.709 but are now displayed in P3 or higher. Skin tones in particular look too red. We will cover this topic in detail in the next DP, especially in the comparison between Windows and Mac.

The PA169CDV – how much colour do you need on the move?

If you want to do something on the go, we recommended the excellent Asus ZenScreen OLED MQ16AH in the penultimate issue – which scored points with its low price and great colours. The PA169CDV (honestly, Asus, we need to talk about names!) is the “professional version”
which is not designed for productivity, but for graphic designers and image makers. What’s in the case?

It’s a portable 15.6-inch screen with 4K resolution (3840×2160 pixels), IPS panel and 10-point touch, which is Display HDR 400 certified. With the included ProArt pen, which uses Wacom EMR technology, you can draw or write directly on the display. Like the PA279CRV, it is also said to have been Calman verified and Pantone validated. Measurement reports are enclosed in each case.

According to Asus, we have 500 cd/m² in HDR mode, 450 cd/m² in typical mode, as well as HDR10 and colour spaces set to Standard, sRGB (100 percent), Adobe RGB, DCI-P3, Rec. 709, Scenery, Reading, HDR and two “User Modes”. In terms of connectivity, there are two USB-C ports on the left-hand side of the screen (one of which is in Display Port Alt Mode) and a full-size HDMI 2.0. Above these are the menu button, an On/Off button and a “rotary wheel”, which can be freely assigned to all kinds of tools. Power is also supplied via USB-C, namely 15 watts in operation and less than 0.3 watts in standby mode. As with the ZenScreen, there is a ¼ inch thread at the rear. Two speakers with one watt each are also installed. The whole thing weighs a little over a kilo, measures 37 by 23.7 by 1.2 cm and costs between €1,100 and €1,300 on Amazon.

However, the smaller PA169CDV costs more – about twice as much for about half the surface area, it is a pen tablet with “Wacom feel”. Otherwise, the screen specs look almost identical at first glance: The currently only supplier for the PA169CDV describes it as 4k OLED but in the same sentence as an IPS panel, which is an LCD monitor type. In fact, both displays are the same: IPS panel with 10 bit and LED backlight
Backlight. And here we are at the first weak point: LED backlights with too few LEDs cause various problems. An old problem with LCD/LED monitors is the lack of black level. This is easy to recognise from the supplied measurement report – it is clearly not an OLED.

The PA279CRV – ProArt for little money!

The second monitor in the test is the Asus ProArt Display PA279CRV Professional 27-inch monitor mentioned at the beginning. A 27-inch 4K (3840×2160) HDR display with LED backlighting, IPS panel and a wide viewing angle of 178 degrees. Asus reports 99 percent DCI-P3 and 99 percent Adobe RGB coverage as the colour space, and that this would be Calman-tested and pre-calibrated to Delta E<2 colour accuracy at the factory.

Everything is available in terms of connectivity, including DisplayPort via USB-C with 96 watts Power Delivery, DisplayPort, HDMI and a USB hub (3.2, and easily accessible at the bottom) plus a wall-mountable design (Vesa 100×100) for maximum desk space. Power consumption is 33.5 watts, the refresh rate is 60 Hz and, because it is currently being discussed: Asus guarantees 8 years of software and firmware updates as well as 7 years of spare parts availability. Very nice, that! But back to the colours…

Calibration right up to the edge…

If you calibrate such a display with the measurement sensor in the middle, the edges/corners will no longer reproduce correct colours. When I first calibrated the PA279CRV and placed it next to my office monitors, I was briefly confused. The monitor actually seemed to have a red cast in the white and I started to check whether I had done something wrong. However, I measured the edges with the measuring probe and measured a considerable red cast, which I also perceived “optically”.

If you place two monitors next to each other and compare them, you focus primarily on both edges and hardly notice that the image actually changes towards the centre of the image, if it actually does. As can be estimated from the amount of area above, the central area (blue-grey cross) is also smaller than the corner areas, which have a different colour. If you take a photo of the white area, it looks clear – I have edited the image for the print.

Incidentally, there is a cool online tool from EIZO for testing monitors: eizo.de/monitortest

I used this to create the white area, you can see the numbers to the left of the centre of the image, if you click on them you can select different test patterns and modify some of them at the bottom right of the window. Anyone familiar with Eizo monitors and their advertising will have seen this picture before. Eizo has developed a technique to eliminate precisely this problem. This is one reason why Eizo monitors are sometimes many times more expensive than the PA279CRV.

Comparison of measurement probe profiles

Here you can see some examples of the profiling problems of measurement probes – this should give you an idea of how important precise individual current profiling of measurement probes is:

Measurement, calibration and evaluation

First of all, it is always interesting to see how different the spectral light properties of monitors from the same manufacturer and of the same or very similar design are. If you use cheap colourimeters, you may have several spectral profiles (usually up to five different light source types such as LEDs or tubes etc.) to choose from, but these cannot be modified, e.g. for so-called phosphor-based LED backlights. However, there are different types such as RG_Phosphor or PFS Phosphor WLED LED backlights with different spectral properties and therefore different white points – if they have not been calibrated to a specific white point. So neither of them can be calibrated correctly with a “standard profile”.

Even if I use a 7,000 euro colourimeter, its stored profiles for perhaps ten specific monitor types such as an EIZO CG series profile are of no use to me at all, as these change with the wear and tear of the years, as with all monitors. This is why built-in measuring probes are more of a consumer gimmick, as they themselves drift and the monitors in which they are installed do too, albeit to varying degrees.

Colours on the PA279CRV

Now let’s take a look at what colours the PA279CRV monitor delivers ex works. Standard colour presets are often “native” and present the entire colour space, which is usually neither fish nor fowl – and always shows the maximum colours possible, i.e. all too often oversaturated.

Dynamic dimming should always be switched off, I had switched to Fast (Display HDR) to see contrast optimised for HDR, but this produces very creepy phenomena, especially in SDR.

Link to the video Dimming-Fail is.gd/dimming

Colour on the 169CDV

If you now start to calibrate the PA169CDV, the gamma unfortunately shifts and the midtones really drop off, a problem that tends to occur with inexpensive colour processing, more expensive devices such as Eizo, for example, have no gamma problems after a white point calibration.

Conclusion

As a long-time Wacom user, I wasn’t particularly impressed with the pen features of the PA169CDV, which is why I haven’t gone into it any further here. With the 16:9 display and the pen aspect ratio, it only fits 16:9 displays such as laptops and desktop monitors. For me personally, the surface is too large when you’re not looking at it and too small to draw on – and too expensive for everyday use due to the limitations. The 169CDV is interesting as a second monitor “to go” on a laptop – sturdily built and the size fits most mobile workstations.

In addition, unlike laptop displays, you can explicitly switch on sRGB or a wide gamut here and work with more usable colour conformity – many laptop displays show “any wide gamut” and you can never be sure what you are seeing. Who says that Apple laptops are very good when it comes to internal monitors? In the next issue, I will go into this topic in more depth and measure and compare an Apple laptop and tablet.

So much for the small one, now for the big one: The large PA279CRV monitor is supposedly great in terms of colour after calibration on the protocol, but you can’t calibrate away the colour casts towards the edge of the picture.

Alternatives!

It’s nice when you’re rummaging around in a range that you occasionally find the cream of the crop. The ROG Strix XG32AQ from Asus, which I bought myself (a quick fix because an office monitor had broken down), is similar in price and larger. This gaming monitor is on a par in terms of overall colour and doesn’t have anywhere near as much colour change towards the edges of the screen for me to recommend it.

Although it “only” has a resolution of 2560×1440, this is much better for working at 32 inches. In general, I find 27 inches with UHD resolution far too small, many menu fonts are barely legible and the accuracy of menu items or buttons with the mouse is drastically less efficient. Other people see it differently, Bela from the DP editorial team thinks 27-inch 4K is the “sweet spot” in terms of resolution and size. But he usually sits much closer to the screen, doesn’t have a tablet and doesn’t click as quickly as I do.

However, if I only want a VideoIO monitor, UHD is preferable – there are no reading problems. But as a video monitor, I would again recommend OLED, with the system’s inherent decent black levels and proper HDR. Because the HDR functions of the 169 and 279 are unfortunately only a “theoretical” feature and hardly usable in real life – 400 nits is only a pro-forma part of the VESA HDR standards, visually HDR 400 is a disappointment if you know what good HDR looks like.

The ZenScreen is an additional monitor for your computer that only requires one cable – and is flexible in terms of form factor and can be positioned very freely – including a cover that can be folded up so that you can easily extend it on the move without straining your back. So far, so understandable.

And anything but a novelty – “small extension screens” have been around for years – but let’s be honest: most of the ones we’ve seen so far are seriously flawed – colour greyness that makes you despair of Excel, energy consumption like the event lighting at a rock festival and manufacturers whose complaints competence is more than dubious. But now that we have one in the magazine, have the devices grown up? Read more!

What is it exactly?

First of all: The “ASUS ZenScreen OLED MQ16AH” (catchy name!) is a portable 15.6-inch OLED monitor (total size including housing 16 inches, weighing 600 grams), which extends the desktop with FullHD (1920 × 1080) and supposedly 100 per cent DCI-P3 – according to Asus with 1 millisecond response time and HDR 10. Mini HDMI and USB-C are available as connections.

And the colours?

The screen is OLED, so the colours are much richer than LCD. But be careful: According to Asus, it is not intended as an “external Class-A monitor”, but for GUIs, light media consumption and all the other things that are normally summarised under the term “productivity”. For those who want more colour fidelity, we will be testing Asus’ “ProArt” screen in the next issue or two – same principle, but designed for colour. But back to the ZenScreen: The various modes in the menu show that it is a flexible helper with no claim to perfect colours. The presets available are: Standard, sRGB, Landscape Mode, Theatre, Game, Night, Reading Mode and the ever-popular Darkroom Mode – it’s clear what that means.

The landscape mode and the theatre mode have extreme gamma and contrast – and are presumably optimised for “outdoors” in order to be able to see anything at all in poor monitor conditions.
But even if it’s not the main application, we couldn’t resist a little measurement – and Calman says that the CIE Illuminant D65 target was hit quite well with a CCT of 6,528 K – with a black point of 0.002 candela, a luminance of 120 candela and a measured contrast of 74,907:1. We measured this with the Calibrite Display Plus HL probe.

The values summarised mean that the screen would be usable for SDR/on-the-go grading (with lightbox) – with the proviso, of course, that transport and lugging around make regular remeasurement necessary. By default, the ZenScreen comes with the standard preset of a P3 Colorspace – so Asus doesn’t want a perfect 100 percent sRGB view, but rather bright colours that make everything more colourful. The panel could be better if it wanted to be, but we are complaining at a high level here.

But in practice..

Let’s be honest: it’s not complicated – just more screen space that interacts with the system in a completely predictable way and is recognised as an additional monitor – and because it doesn’t make any problems and offers solid colours, it quickly became the dedicated “near” screen in everyday editorial work, while “communication” and browsers were running on the other screens – the familiarisation time was minutes. For video processing (in this case with Premiere), we used the entire additional area as a transcription window – and in one case as a waveform area.

As you can see in the screenshot, the screen (attached to a small tripod) was directly above half of the keyboard – while the main screen is directly above it. As both brightness and contrast are excellent for everyday use, this feels very much like a classic typewriter – especially if you’re not typing completely blind. Other applications are possible! Thanks to the tripod thread and the low weight (650 grams), microphone arms and the like are also possible – if you want to have it at a specific workstation position. And of course, the standard thread means that there are cheap accessories for every situation – be it clamps, stands, holding straps or anything else.

Screen in a box

A very clever idea from the developers was to make the box multifunctional. The transport “case” made of sturdy, black cardboard with a few small magnets can be folded into a “light protection bonnet”, which makes a good picture on the DIT trolley or during a streaming job, for example – or provides a certain amount of privacy in other applications. Four “foam corners” are included for fastening, and there is plenty of space under the screen for cables and accessories – or even the emergency flat man (we’ve all had SUCH productions!) really everything in a small box. By the way: If you often use the screen when travelling, you can also order a screen protector in the exact dimensions (see here: is.gd/amazon_schutzfolie_asus).

Pro

What we like about the ZenScreen: A great deal – for small, light, flexible, surprisingly usable panel and it fits into the workday without bitching as long as you have a USB 3/USB-C port. With its low power consumption, it won’t drain your laptop battery, and “firmly” set up and positioned, it noticeably increases efficiency. And, also cool, is the box – a sturdy mount that’s big enough for two people to see. Although it’s not as sturdy as a dedicated photo viewing cover, Asus cleverly avoids packaging waste – the box is sturdy enough for occasional use, and if you throw it away, it’s really your own fault. There is also a “proximity sensor” – so if the device is attached to the laptop but you walk away, the screen won’t drain the battery.

Cons

Sooo, but let’s get down to the niggles: It’s just another device on the table – and if you want to set it up as “cleanly” as possible, you might want to get an angled USB cable. And the second gripe? The stand cover has exhausted our (admittedly extremely rudimentary) origami skills – and we simply used a tablet stand for “permanent use”. And then after a few days we replaced it with another “mini stand”, which meant that the height was ideal. So: neither of these complaints are deal-breakers, but we just wanted to say so.

Conclusion

So, to summarise, who is the ZenScreen suitable for? In our opinion, for anyone who needs more screen space but doesn’t have room for a “full” monitor – as well as for those who want to work more efficiently on the move or have jobs where they want to use more complicated software. We can definitely recommend it – the average price on Amazon is 450 euros, used/refurbished devices are also often available at around 350 euros.

At Asus: is.gd/asus_zenscreen_mq16ah

Display size: 15.6 inch
Aspect ratio: 16:9
Visible area: 344.21 x 193.62 mm
Panel: OLED, 10-bit
Viewing angle (CR≥10, H/V): 178°/178°
Resolution: 1920 x 1080
Colour space (DCI-P3): 100 percent
Brightness: 360 cd
Contrast (HDR, Max): 1,000,000:1
Contrast (Typical): 100.000:1
Response Time: 1 ms (GTG)
Refresh Rate (Max): Typically 60 Hz
HDR Support: HDR10

Video Feature
Trace Free: Yes
Colour Temperature Selection: Yes (4 Modes)
Colour Accuracy: ∆E≤2
GamePlus: Yes
QuickFit: Yes
(Photo/Alignment Grid)
HDCP: Yes, 1.4
Dark Boost: Yes
DisplayWidget: Yes,
DisplayWidget Lite
Low Blue Light: Yes

I/O
Two USB-C ports (DP Alt Mode)
Mini HDMI
Digital Signal Frequency:
HDMI: 60 HZ (V) / USB-C: 60 HZ (V)

Power consumption
Consumption: 4.84 W
Standby: < 0.5 W
Voltage: 100-240 V, 50/60 Hz

Dimensions
Without stand (W x H x D):
359 by 227 by 9 mm
Lightbox (W x H x D):
550 by 390 by 125 mm
Weight: 650 grams
Thread: 1/4″ tripod thread

Accessories
Mini-HDMI to HDMI cable
Power plug
Quick start guide
USB type C to A adapter
USB-C cable
Tripod thread cover
ZenScreen “Smart Cover”

This article originally appeared in Digital Production 02 : 2014.

In my article “4K & HFR” in DP O7/13, I covered the technical basics on the subject of “True 4K”. Now it’s time to put it into practice.

To get you started, here is a summary of the most important facts:

• • A 4K monitor/projector has 8.85 megapixels at 4,096 × 2,160 according to the DCI standard. As we know, a pixel consists of three colour pixels: red, green and blue. This means that 8.85 megapixels times three equals 26.5 megapixels in RGB.
  • However, a camera sensor labelled as 4K only has a total of 8.85 megapixels for RGB for 4,096 × 2,160, and only 2,048 for green on the horizontal. However, it would actually have to have 26.5 megapixels to achieve the same resolution as the monitor. A camera sensor currently labelled as 4K therefore only achieves a luminance resolution of 2K! Tip: Just google Bayer pattern sensor.
  • To deliver true, full 4K, i.e. 26.5 megapixels for RGB, you would need an 8K×4K sensor as a minimum, which neither the Epic Dragon sensor nor the Sony F65 currently achieves.
  • In order to be able to see real 4K (which currently only exists in the photo and animation sector), you must not be further than 1.5 times the image height away from the image.
  • In order to be able to perceive real 4K while the image content is moving, you need an image motion resolution/frame rate above the perception threshold of around 90 to 100 frames per second (HFR = high frame rate)

With this in mind, it is fair to ask: does it make sense to even try to produce 4K at the moment?

Projection

The good news is that there are actually applications where 4K makes sense even with the current technical framework conditions: for large projections at events (example: car shows). Anywhere where viewers (e.g. visitors to trade fair stands) can get close to the screens. Or in the field of design and development, where only artificially digitally generated images are presented, sometimes also many still images or paused virtual live 3D animations.

With computer-generated images, there is usually no limit to the frame rates, as long as the computing power is sufficient. The automotive industry has been working with 4K projections under these “conditions” for around ten years.

There is another aspect to large projections: Even if you can no longer perceive 4K from a distance of 1.5 times the image height to the screen, the pixel structure can still be perceived from a distance of 10 times the image height with a 2K or HD projection, especially with high-contrast graphics or fonts. A 4K projection, even of HD content, would therefore look much better in the last cinema row, because the scalers built into 4K projectors convert this pixel structure into smooth lines. Or if you produce the graphics in 4K and the film content in HD, it looks better on the 4K projector than on an HD projector.

Alternative applications

Another application variant for 4K was presented by Canon in its latest roadshow: Photo shoots with the Canon 1D in 4K video mode. Here at least 24 images per second can be recorded in 4K (but only in Motion JPEG, quite heavily compressed with limited dynamic range compared to raw photos).

The photographer then no longer presents printed photos, but short snippets of movement on 4K monitors with a long pause in between – sometimes just a blink of the eye or a twitch of the corner of the mouth. A completely new art form? Well, not quite. Basically a video installation, albeit in unprecedented image quality, which certainly fulfils the increased image quality demands of viewers.

For consumers, there are also some areas of application for 4K that they can already utilise. Viewing photos: Even an iPhone takes photos with 8 megapixels, as many pixels as a Sony F5 or F55 4K sensor has. A photo from a Canon 5d has over 20 megapixels, a Nikon D800 has 36 megapixels – such photos look simply fantastic on 4K displays. You can see the full quantum leap of 4×HD.

If you want to achieve this quality in 4K moving images: timelapses, i.e. series of individual images from such cameras, are the only true 4K content to date, apart from computer animations. Current new games consoles or games computers can also reproduce 4K up to 60 FPS.

However, there are also other examples. Sony F65 users tell us: “We bought an F65 because we saw that a Red or an Alexa reach their limits on the huge screens that Audi operates at these trade fairs. The images just look soft.” (The quote comes from the application report of a 4K production with the Sony F65, in which the frame rate problem is also vividly described: Kropac-Media at http://bit.ly/1dqNb9f.)

Apart from timelapses, maximum resolution in images can only be achieved digitally and artificially, or you have to come up with multiple resolutions and scale very carefully. However, there is still a lack of cameras that can produce true 4K and, above all, those that significantly exceed true 4K resolution (26 megapixels plus!) and do so at a simultaneous frame rate of at least 60 FPS. The most common argument in favour of 4K productions that I have heard recently was: “Then we can still zoom into the image because we have so much resolution.” Or: “We only shoot the slow motion in 2K, that’s still enough for HD utilisation.” However, if you record 2K cropped with a Bayer pattern sensor, you are left with significantly less resolution than HD. 2KBayerpattern resolution is just about good enough for 16:9 SD.

HFR – more than 30 FPS required

Another quote from the aforementioned application report reads: “In our tests, we realised that 4K resolution is only half the battle. The essential thing is recording in 50p. This is what creates the real wow effect. Only then does the material look really sharp.”

HD at 59.94 frames per second has been the standard for BMW trade fair films for over ten years. Audi now also requires every large-scale projection in 4K to be delivered in at least 50p, both filmed content and 2D/3D animations. For cinema projection, the frame rate in the DCI standards has been increased to up to 60 frames per second, particularly for stereo 3D. Almost all high-end finishing systems (such as those from Quantel and DVS) can also process at least 2K to 60 FPS in stereo 3D or 4K 2D at 60 FPS.

However, consumer TV sets and computer displays pose a problem when it comes to 60 FPS. I recently visited a production company that is currently producing films in 4K for a TV set manufacturer for product presentations of 4K TV sets. This visit led to a joint field research odyssey, which I would like to report on below because it clearly illustrates the “teething troubles” of 4K.

Great expectations and unvarnished reality

With its 65 inches, the TV set supplied by the manufacturer for testing already offered an impressive picture size, creating anticipation of brilliant images. In addition, the playback device from the same manufacturer – a kind of tablet with a docked keyboard – was supposed to make 4K easy and simple for anyone to operate with the swipe of a finger, so to speak. So much for the marketing. And that was the reality: after several attempts to connect and disconnect the 4KTV device to various HDMI ports on the computer (Windows 8.1), it recognised itself with 4K, i.e. 4,096 × 2,160. When the first images lit up, the colleague who had previously dragged the heavy device into the office by the sweat of his brow couldn’t contain his excitement: It looks great! Sitting on the sofa about three metres away, he didn’t notice the image deficiencies or the lack of sharpness. About one metre away from the screen, however, the unvarnished 4K reality was revealed in the form of massive image resolution losses.

Massive loss of image resolution

Now we began our search for the causes of the disappointing picture quality when viewed up close. To anticipate, at the end of this search, which can certainly be described as an odyssey, we were enriched by the following findings:

• • • Monitor sends incorrect resolution information to graphics card.
    • Graphics card scales up Quad HD to DCI standard (because the Quad HD film is displayed in full screen).
    • Then the monitor scales down from DCI standard back to its native Quad HD resolution and zooms into the picture using overscan.

Here is our field research odyssey in detail:

• • Check: Resolution / image content

As we didn’t have a 4K test chart, we produced one: First, we downloaded my HD test chart from digitalproduction.com (http://bit.ly/1b2tUXr) and created a Quad HD, i.e. 3,840 × 2,160, in After Effects and then also in 4,096 x 2,160 by filling it up. The aim: to determine whether the monitor has DCI standard or Quad HD.

• • Check: Resolution/picture reproduction devices/playback system

Loading the test chart revealed that a resolution below HD could already be seen in the result when this still image was played back. So we looked in the graphics card settings to see what was offered as an alternative. The next entry was 3,840 × 2,160 – but without the additional information “native”, which was also missing for 4,096 × 2,160. This is often the case with computer displays so that it is easier to set the appropriate native resolution or it is initially used automatically. However, the image only became a little sharper, it was still not pixel native.

• • Check: Resolution/picture display device/TV set

We searched for the scaling or overscan in the display menu, as there is no standard designation here and it can be called “full”, “1:1”, “native” and similar. When we found the setting, it still looked blurred, but pixel-perfect. My test chart has various elements that can be used to quickly differentiate such a problem.

• • Check: Image processing functions/ TV set

• TV set manufacturers apparently assume that the majority of picture content played back to consumers at home is of such poor picture quality that the TV sets process the picture quality with a considerable number of functions. However, these picture enhancement functions prove to be counterproductive for high-quality picture material. It is therefore advisable to switch off or correctly set all these functions, which affect saturation, colour space, colour temperature, frame rate algorithms and, in particular, sharpness. Normally, sharpening is always active and inactive at a value of 0, but on our device it turned out that image sharpening was inactive at 50 (the default value). I was only familiar with this type of setting from cameras that can also go into the negative, i.e. can even blur the image. After I had deactivated the image sharpening function, the test chart finally looked perfect and pixel-perfect.

• • Check: Frame rate/image display devices

• In the graphics card settings, the frame rates were hidden in the Advanced Settings. After we had changed the resolution from 409 2,160 to 3,840 × 2,160, we were now able to set 25 FPS, 29.97 and 30 FPS instead of just 23.97 or 24 FPS. As the 4K film had been produced at 25 FPS, we now set 25 FPS accordingly. This showed that the higher frame rates of up to 30 FPS could also be activated at 3,840 × 2,160 – a further indication that this was the native picture resolution of the TV set’s panel. To be able to display 4,096 × 2,160, the internal scaler has to utilise additional computing capacity, which is at the expense of the frame rates. In the end, we exported the 4K file at a significantly lower data rate, i.e. from the original 300 megabits to 25 megabits. Fortunately, the production company had produced the file in Quad HD, i.e. in 3,840 × 2,160, which now also corresponded to the resolution of the TV set. After all these tedious changes to the settings, the 4K film looked halfway decent if you were standing two metres or more away from the screen.

• Final check: Motion resolution

Now that the TV set was also displaying the native frame rate of the film and we were able to play the H.264 film with the now significantly lower data rate of 25 Mbit almost smoothly, the differences in quality from almost motionless image content to the faster ones were already drastically visible if you stood close enough to the 4K TV set. There are at least two reasons for this fluctuating picture resolution quality:

a. Sensor resolution for slow motion: In order to achieve higher frame rates, the resolutions of the camera sensor usually have to be cropped. This means that the available pixel quantity is not fully utilised – often in order to save storage capacity. For example, even the Sony F65 only achieves 4K × 1K at 120 FPS, i.e. only a quarter of the resolution. However, the sensor has 8K × 2K.

b. Motion blur due to excessively long exposure times: Motion blur. The higher the actual still image resolution and the closer the viewer looks at the image on higher-resolution image devices, the more noticeable the differences in image resolution. Arri has already shown in various presentations impressive tests how dramatic the image resolution losses are with 4K due to motion blur. It therefore often makes more sense to produce consistently in HD or 2K with at least 60 FPS than 4K with 24 FPS. Especially when 4K cannot be real 4K.

HDMI 2.0 must come soon

At least the TV set in our field research had an HDMI interface of standard 1.4, which is the minimum required for 4K. The computer/playback device also had HDMI 1.4, although HDMI 1.4 is only specified up to 30 FPS. But the frame rate was automatically selected by the graphics card/TV set: 24 FPS. An average consumer would already be overwhelmed by the task of setting the device correctly.

Only HDMI 2.0 will support 60 FPS. Until the end devices can handle this, 60 FPS films should not be released. However, they can be produced, as various consumer TV manufacturers are already pointing out that their devices will be easily upgradeable to HDMI 2.0. “Easy” because the HDMI interfaces are often built into docked boxes to enable such upgrades. You can already produce at 48 or 50 or 60 – and then simply output half the images.

For example, “The Hobbit” was shot at 48 and not 60 in order to be able to deliver a normal 24 FPS for 2D cinema exploitation, which doesn’t shutter so violently. With 60 FPS, the film would have been shot much faster and the exposure would have been shorter, which would have led to more shutter problems. So if you want to be downwards compatible, you always have to shoot in such a way that you only have to halve the FPS. So 60 FPS for a 30 FPS output or later 60 FPS.

With HDMI 2.0, 30 FPS can also be displayed with up to 12-bit colour depth, making a higher colour resolution than in post-production possible for the first time. However, only the previous 8-bit colour depth will be possible again at 60 FPS. Either way, the amount of data and therefore the technical requirements will increase dramatically.

Bear in mind that wireless HDMI transmission only works up to Full HD. Wireless 4K will not be possible for a long time yet. The requirements for cable lengths and their quality will also increase considerably with 4K. Especially if instead of 30 FPS, 60 FPS, i.e. twice or even almost three times the 24 FPS data rate, has to be delivered via the cable.

Tip: To solve such a problem in production, you can still fall back on classic SDI technology. There are various affordable products in the adapter sector from Blackmagic, AJA etc. The first 4K live productions provide corresponding reports of experience: we will still have to work with HD technology for a long time to come, especially when monitoring 4K productions, as the effort and costs would otherwise explode: Videos and additional material from a live music production: http://www.4k-concerts.com

Conclusion

Even if the sensor/camera and TV set have been optimally selected/adjusted, the decisive factor for achieving true 4K quality is the production. In the case of our 4K TV device field research, the content was produced with Red Epic in Quad HD and with Sony F700, whose sensors are read out cropped in Quad HD. So only with just under 8.3 megapixels in RGB. A Quad HD TV set has 24.9 megapixels, i.e. around three times the resolution in RGB and around twice the resolution in luminance on the horizontal plane. Accordingly, graphics/fonts/logos are crisp and sharp, but the “4K film content” behind them is still quite soft, although no longer pixelated. The majority of readers probably know from their VFX experience that they sometimes have to degrade the artificially generated, crisply sharp graphic content by means of grain, noise, chroma aberrations and blurring so that it harmonises with the “real shots”. Well, this additional workload will also remain with 4K. I dread the thought of having to accommodate SD archive material in 4K.