News
Buyer's Guide: The Best Monitors 
Articles >
Digital Camera Screen Tests: First Results, First Conclusions
Last month, we developed a procedure for testing the screens built into digital cameras. Since then, we've put 17 camera screens to the test with our special sensor and we've seen some very mixed results. Here's a brief overview of what we've found and what preliminary conclusions can be drawn from the results.
Franck Mée
Translator: Catherine Barraclough
Updated: January 4, 2011
Translator: Catherine Barraclough
Updated: January 4, 2011
Delta E
Underneath its rather complicated-looking name, Delta E actually quantifies a fairly simple concept: the difference between a colour a screen is requested to display (by a photo file, for example) and the colour it actually displays. It's sometimes known as the 'colour difference', and basically measures how far the screen is off reproducing an ideal value. A perfect screen would have a Delta E of zero for all colours. If the Delta E is under 3, a colour can be considered to be reproduced accurately as the difference won't be visible to the naked eye. When the average Delta E of all the colours measured is under 3, the screen is considered to be accurate.
This however, is very rarely seen in digital camera screens, where the Delta E is usually in the realms of 7 or 8, which would be quite poor for a laptop or monitor. Some camera screens have a Delta E that easily reaches 10, and which is really quite bad. On that kind of screen, the colours displayed will only give an approximation of those actually captured and saved in the photo file. Such screens should definitely not be used for sorting through and deleting photos, as they're simply not trustworthy.
This however, is very rarely seen in digital camera screens, where the Delta E is usually in the realms of 7 or 8, which would be quite poor for a laptop or monitor. Some camera screens have a Delta E that easily reaches 10, and which is really quite bad. On that kind of screen, the colours displayed will only give an approximation of those actually captured and saved in the photo file. Such screens should definitely not be used for sorting through and deleting photos, as they're simply not trustworthy.
However, strangely enough, it's been considered pretty much normal for digital cameras to have screens of approximative quality. This is often justified by the fact that camera screens are viewed in all kinds of different conditions, from a dark nightclub to outdoors on a sunny day, and that these changes in lighting will have too big an impact on how our eyes interpret what's on the screen.
.jpg)
But many users have found that a photo taken in their living room looks very different when viewed on a monitor in the very same room: it can sometimes look like a completely different picture! Have you ever showed your friends a photo with bright blue skies and lush green trees on the camera screen, then transferred it to a computer to find it pale, washed-out and dull?
Plus, at a time when more and more cameras are offering instant uploads to sites such as Facebook and Flickr, the autumnal tree tinged with red that looks fantastic on your camera screen could end up just looking slightly yellow to your friends online. The original appeal of the picture is thus completely lost as the camera hasn't captured what you thought it had!
So the main advantage of an accurate camera screen is the peace of mind that the photo you view on the camera will be exactly the same as the one stored on the memory card, the one viewed on a computer monitor or the print you get developed.
We therefore decided to put together a test procedure to assess the accuracy of the screens used in digital cameras. We started including this in our reviews about a month ago, and we've currently tested 17 cameras with some highly variable results.
In this article, you'll find an overview of all of the results from our camera screen tests to date, with explanatory diagrams like those you may be used to seeing in the monitor face-off. Those of you who want to skip the list of results can go straight to our analysis of these findings, where we'll try to draw some preliminary conclusions from what we've seen. Note that any conclusions drawn here are only based on the 17 tests carried out to date—we're still hoping they'll be proved wrong at some point in the future when manufacturers start using perfect, or at least acceptable, screen calibration.
Results
From left to right, the three diagrams next to each camera show:- gamma, or the accuracy of the grey scale (on average, are the greys too dark, too light or well balanced?);
- colour temperature, from dark to light shades (does the on-screen image look too warm and have a yellow tinge, or does it look too cold and have blue overtone?);
- Delta E, or the accuracy of the colours displayed (see sidebar).
Consumer Compacts
Delta E: 15.7 (awful)
Gamma: 1.4
Colour temperature: 9100 K (poor)
Delta E: 11.3 (bad)
Gamma: 1.4 (highly irregular)
Colour temperature: 7200 K
Delta E: 10.7 (bad)
Gamma: 1.5 (highly irregular)
Colour temperature: 6600 K
Delta E: 11.5 (bad)
Gamma: 1.7 (highly irregular)
Colour temperature: 7000 K
Delta E: 7 (acceptable)
Gamma: 1.7
Colour temperature: 7000 K
Delta E: 8.7 (bad)
Gamma: 1.7 (highly irregular)
Colour temperature: 10300 K (poor)
Delta E: 1.6 (bad)
Gamma: 1.5 (highly irregular)
Colour temperature: 8800 K
Delta E: 8.8 (bad)
Gamma: 1.5 (highly irregular)
Colour temperature: 7100 K
Expert Compacts
Delta E: 8.4 (bad)
Gamma: 1.7 (highly irregular)
Colour temperature: 7900 K
Delta E: 4.4 (bad)
Gamma: 1.9 (irregular)
Colour temperature: 5600 K
Consumer SLR/Bridge
Delta E: 7 (acceptable)
Gamma: 1.7
Colour temperature: 7000 K
Delta E : 4.7 (not great)
Gamma: 1.9 (highly irregular)
Colour temperature: 6500 K
Delta E: 5.7 (not great)
Gamma: 2.2
Colour temperature: 7100 K
Expert SLRs
Delta E: 2.4 (good)
Gamma: 2.2
Colour temperature: 6800 K
Delta E: 2.2 (good)
Gamma: 2.1
Colour temperature: 6700 K
Delta E: 5.6 (not great)
Gamma: 2.3
Colour temperature: 6000 K
Professional SLR
Delta E: 1.7 (very good)
Gamma: 2.2
Colour temperature: 6600 K
Grey Scale: Room for Improvement
To be honest, when we developed our test procedure for digital camera screens, we weren't really sure what kind of results we'd see. However, from years of hands-on experience, we knew that most compact camera screens would be quite poor quality. We also suspected that the cheapest models would be those with the worst screens, as they're obviously equipped with the lowest-quality panels, often sourced from a wide range of different suppliers. It therefore wouldn't be cost effective to develop a tailored colour management system for each variant.However, some of the results were more surprising. For example, it seems that the colour temperature is now stable in most cameras, correcting a common complaint of desktop monitors just a few years ago (colours looked too blue in dark zones and too red in light zones). So although the average colour temperature varies from one model to the next and is generally a little high (between 7000 and 8000 K), it's now remains more-or-less stable no matter what the brightness. Exceptions to this include the Sigmatek SDC-350p, a very low-end camera that probably has an old-generation screen and, more surprisingly, the Panasonic TZ7, in spite of its high-end status.
Big improvements still need to be made to the gamma, however. Although the gamma is generally OK in SLRs, all the compacts we've tested to date have a very irregular grey scale—light greys, in particular, turn almost white very quickly. The exception to the rule is the Nikon S5100 (review coming next week), which has a low but fairly regular gamma of 1.7. In other words, the mid-range greys are too light but the levels of grey are clearly distinguishable and the difference between them is constant.
Colour Reproduction: A Lottery
Delta E seems to be rather more predictable, with results that improve consistently as you move up camera ranges. In other words, colour reproduction is all over the place in entry-level cameras but a bit more effort is made with 'expert' models, with at least a basic level of colour management. Things generally get better still with entry-level SLRs, even though certain colours are accentuated to give more striking but less accurate pictures.The only really well-calibrated screen we've seen so far is in the medium-format Pentax 645D professional SLR—a camera that'll set you back the best part of £1,000. The funny thing is that the screen used in this model is actually identical to the one used in the Pentax K-5, which got the worst results in these tests out of all three expert SLRs released in autumn 2010, with highly inaccurate blues and greens. We therefore reckon that the Pentax 645D has been properly calibrated after production and before sale. At this level, cameras could perhaps even be calibrated on a camera by camera basis. The K-5, on the other hand, has been calibrated for the consumer market and may actually have been left with the screen's default factory settings.
Range and Price
From what we've seen so far, the general rule seems to be that for under £300 you'll get a low-quality screen or a good-quality but ill-calibrated screen. The Nikon Coolpix S80, for example, has an OLED panel with a gamut so wide it can almost display fluorescent colours. However, it doesn't handle this wide gamut well and in turn over-saturates pictures with almost comical consequences.Expert camera ranges tend to display colours more accurately, although the grey scale can still be hit-and-miss. Cameras with interchangeable lenses generally have more accurate displays than compact cameras, and the quality seems to increase as you move up a range towards the higher end. One notable exception is the Panasonic GH2, with a screen that overexposes light grey tones.
The quest continues!
> Camera Reviews
