Take your photography to the next level and beyond...

  • NEWS
  • REVIEWS
  • INSPIRATION
  • COMMUNITY
  • COMPETITIONS

Why not join for free today?

Join for Free

Your total photography experience starts here


Nikon D90, Canon EOS 550D & Pentax K7 Comparison Test Digital SLR Review

Nikon D90, Canon EOS 550D & Pentax K7 Comparison Test Digital SLR Review - Duncan Evans puts the Nikon D90, Canon EOS 550D and the Pentax K-7 through their paces in a head to head comparison test.

 Add Comment

Category : Digital SLRs
Product : The great DSLR shootout
Share :





v




v
Canon EOS 550D with 18-55mm lens - available from Warehouse Express for £643.00 Nikon D90 body only - available from Warehouse Express for £609 (the 18-55mm lens costs £147) Pentax K-7 with 18-55mm lens - available from Warehouse Express for £788.00.

Test 1: JPEG v RAW
Test 2: ISO speeds
Test 3: Noise reduction
Test 4: Sun in the frame
Test 5: Backlit, contrasty scene
Test 6: Bright sky, dark foreground
Test 7: Predominantly black subject
Test 8: Predominantly white subject
Test 9: Continuous focus and fast shooting
Test 10: AWB in daylight
Test 11: AWB in cloudy conditions
Test 12: AWB under tungsten lighting
Test 13: AWB under mixed fluorescent lighting
Test 14: AWB using flash
Test 15: Custom WB
Test 16: 18mm focal length
Test 17: 35mm focal length
Test 18: 55mm focal length
The_conclusion

All three cameras have previously been tested on ePHOTOzine but tested in isolation is one thing, what about when they face the same conditions at the same time in a very practical test, rather than in the lab? It’s at this point you start to get some idea about which camera is best at doing what function. Whether that’s the quality of the lens, the ISO noise reduction or ability to deal with difficult exposures. In this shootout then, all the cameras were tested in the same conditions at the same time, shooting RAW for maximum quality unless stated otherwise. The Canon was equipped with a EF-S 18-55mm f/3.5 IS lens, the Nikon with a Nikkor 18-55mm DX VR f/3.5 DX VR lens and the Pentax with an smc Pentax-DA 18-55mm AL II lens.

Firstly though, a word about the conditions involved. All digital cameras process their images – otherwise there wouldn’t be any. This isn’t the age of film where the same film in different cameras can be directly compared. So, to get as close to raw picture quality as possible, all the highlight and shadow enhancement, colour and contrast adjustments and sharpness options were turned off in the firmware. Even the high ISO noise processing was turned off, except where it was specifically being tested. When looking at the RAW files in Photoshop, all the import filters that are normally applied when loading were set to zero. Each camera used the sRGB colour profile, the neutral or normal colour and contrast profile with everything set to zero enhancement. The Canon RAW files were CR2, the Pentax PEF and the Nikon NEF.

Canon EOS 550D, Nikon D90, Pentax K-7: Image quality

Test 1: JPEG v RAW
For this test the cameras were set to shoot RAW and top quality JPEG at the same time. It’s important to realise that the camera isn’t taking two shots here, it’s taking one, saving the RAW file as the basic file and the JPEG one after applying a set of specific algorithms. The advantage that RAW has is that you can set the processing priorities later. When loading the RAW file to compare against the JPEG, they look very different, but that’s because the RAW import contains a set of parameters for hue, brightness in particular, contrast, sharpness etc. Reset all these import settings to zero and the difference between the RAW and JPEG file is dramatically reduced. So, your results in practice will depend on exactly how you decide to process the RAW file. If you don’t want to, then use the JPEG files, learn your cameras tendencies and adjust your shooting parameters accordingly.

Now let’s get to the comparisons. In the case of the Canon, the JPEG image is brightened up a lot to the point that the brickwork on the left was overexposed. With no adjustments to the RAW file on importing there’s a much better result with more detail on the walls. In terms of detail, it’s very close, there not much lost at all in the high quality JPEG save. Sharpness as well, is almost the same, but the CR2 is a little sharper. The tone of the RAW file is a little warmer and obviously, in this example, the colours are better because it now isn’t overexposed. However, colour fringing on the RAW is certainly more apparent than after the little tweak the JPEG file gets. The other thing is that shadow noise is ever so slightly more apparent in the RAW file because basically it’s been smoothed out in the JPEG file.

To the Nikon then, and the first point to note is that the JPEG file is warmer than the NEF file which is slightly cool in comparison. Again though, the JPEG file is much brighter, though the brickwork isn’t as overexposed as the Canon JPEG. Also, you can see more noticeable colour fringing in the RAW file than the JPEG. There’s slightly more contrast in the JPEG file (even though it’s brighter), it retains as much detail as the RAW file but is softer.

Progressing to the Pentax the PEF RAW file is underexposed. There’s clearly space on the right hand side of the histogram for more highlights. However, this has given it the perfect opportunity to brighten the image for the JPEG which is spot on. The RAW file is clearly sharper than JPEG file which also loses a little detail. Tonally, they are much the same, bearing in mind the brightness difference.

Direct comparisons are interesting because the Canon metering has gone for a brighter image compared to the Pentax and Nikon. Shooting at f/8 in AP mode with centre-weighted average metering in all cases, the Pentax and Nikon produced shutter speeds of 1/200th sec, the Canon 1/125th sec. The extra light certainly brightens up the shadows, but the JPEG processing of the Canon has distinctly let it down by wading in and brightening it up to the extent that the brickwork is completely overexposed. The Nikon looks like it brightens just as much, but starting from a lower light level the result is better. The best outright result from an exposure perspective came from the Pentax which underexposed the RAW file, but brightened it and added just the right amount of contrast to give a near perfect JPEG file.

Now, stacking the JPEGs up against each other, the Nikon file is the sharpest with more detail. The Pentax file has the most contrast, but also slightly more tonal variation in the shadows. The Canon file is the softest with the least detail, but bear in mind, the differences are very minor.

Comparing the RAW files directly and the results are very similar but these observations can be made. Firstly, that the Pentax file is the sharpest, that the Canon is the softest, but they are extremely close. The Pentax has the most contrast but also more noise in the shadows than the other two. And now I’m going to add the caveat that these are the results obtained with the lenses used for the test. Different lenses will give different sharpness and contrast, but they won’t affect the noise or how the firmware treats processing the JPEG from the RAW data.

The great DSLR shootout JPEG v RAW: Click on the thumbnails for the larger images.
  Canon EOS 550D Nikon D90 Pentax K-7
JPEG Canon EOS 550D JPEG Nikon D90 JPEG Pentax K7 JPEG
RAW Canon 550D RAW Nikon D90 RAW Pentax K7 RAW

Test 2: ISO speeds
All the cameras were tested through the range at which they can be compared, which is from ISO100-6400. For this, the Nikon has Low and High ratings, but they have the same speeds as ISO100 and 6400 respectively. The Canon also has a High mode of ISO12,800 which is very, very noisy. All camera and import filter based high ISO noise reduction was turned off to see the basic picture. So to the ISO100 setting and it’s good news all round. There’s no noise here for any of them. At ISO200 there’s some tonal variation in the shadows of the Pentax, but then this has more contrast so it’s going to make it more apparent. The Canon looks very slightly more gritty the Nikon is much the same as at ISO100. At ISO400 the Nikon and Canon are still performing very well, there’s only very minor effects, but the Pentax now shows clear tonal variation in the shadows, it’s clearly worse than the other two, but in itself, doesn’t affect the overall picture quality. At ISO800 both the Nikon and Canon show variation in the shadows and the Canon has some colour variation in the stonework. The Pentax has noise now in the shadows and it’s also apparent across the leaves and other areas of the image.

The fun starts at ISO1600 with much more clearly defined, coloured noise across the entire Pentax image. The Canon also now has coloured noise across the image, though it’s softer, while the Nikon shows better quality in the lighter areas, but the same kind of noise in the shadows. Very noisy shadows for the Pentax at ISO3200 and hard noise across the main image. The Nikon also has clumps of noise in the shadows and soft noise in the general image but it’s significantly better. The Canon is also better, with soft noise. At ISO6400 there’s sharp coloured noise all over the Pentax image with loss of detail in the shadows and the image is certainly darker. The Canon has lots of coloured noise across the image, though as before, it’s certainly softer than that in the Nikon image. The Nikon has blotchy noise in the shadows and all over the image, but it’s not too bad. Overall, this is the one area where the Pentax does fall behind the Nikon and Canon. Once you get past ISO200 it develops more noise at each setting. At the high end, the Nikon tends to sharper clumps while the Canon is softer but across the entire image. How noise reduction works at these high numbers though, is the question answered in the next test.

The great DSLR shootout ISO speed test: Click on the thumbnails for the larger images.
ISO Canon EOS 550D
Nikon D90
Pentax K-7


100
Canon EOS 550D ISO100 Nikon D90 ISO100
Pentax K-7 ISO100


200
Canon EOS 550D ISO200 Nikon D90 ISO200 Pentax K-7 ISO200


400
Canon EOS 550D ISO400 Nikon D90 ISO400 Pentax K-7 ISO400


800
Canon EOS 550D ISO800 Nikon D90 ISO800 Pentax K-7 ISO800


1600
Canon EOS 550D ISO1600 Nikon D90 ISO1600 Pentax K-7 ISO1600


3200
Canon EOS 550D ISO3200 Nikon D90 ISO3200 Pentax K-7 ISO3200


6400
Canon EOS 550D ISO6400 Nikon D90 ISO6400 Pentax K-7 ISO6400

Test 3: Noise reduction
Although the options are well buried in most cases, there’s a choice of noise reduction at high ISO ratings for all three cameras. The Canon has two settings, with the default Strong setting being automatically applied unless it’s turned off. The Nikon has four strength settings, but like the Canon, there’s very little difference between them so you may as well use the highest setting. The Pentax has five settings and there’s quite a big difference between them, from hardly any effect to much more aggressive reduction. Stacking all three up against each other at ISO3200 on the strongest setting, there’s a case to be made that the noise reduction of the Pentax is the best because it has the worst noise and yet it makes a good job of it. The Canon noise is softer than the Nikon but is more prevalent. The Nikon reduces it to clumps so some areas look better and others worse. From a straightforward question over which image looks best at ISO3200 with the noise reduction set to maximum the answer would be the Nikon, though there’s not much in it.

The great DSLR shootout noise reduction test: Click on the thumbnails for the larger images.
Canon EOS 550D Noise Reduction Nikon D90 Noise Reduction. Pentax K7 Noise Reduction
Canon EOS 550D Nikon D90 Pentax K-7

Canon EOS 550D, Nikon D90, Pentax K-7: Awkward exposure situations

Test 4: Sun in the frame
Once the sun starts to intrude on a shot you can guarantee that the metering will start to panic. It introduces an element that is far brighter than anything else in the scene, but the question is will it overcompensate? The zone metering system for each camera was used so as well as the bright sun they should take into consideration the scenery and sky. The darkest image is the one from the Pentax, which also explains why the burnout from the sun is smallest. The Canon is very similar to the Pentax but in straight zone metering terms the best result is from the Nikon which has weighted the scene with more consideration of the other elements, giving a better overall exposure. It’s also come up with a bluer sky.

The great DSLR shootout sun in the frame test: Click on the thumbnails for the larger images.
Canon 550D sun in the frame Nikon D90 sun in the frame Pentax K7 sun in the frame
Canon EOS 550D Nikon D90 Pentax K-7

Test 5: Backlit, contrasty scene
For this next test, the scene is backlit. The sun in the clouds behind the cathedral making the clouds very white. The foreground isn’t dark as such, but it does contrast with what’s in the background. The important point to note is that the vast majority of areas in this scene are darker than the sky. Both centre-weighted and zone metering were tried here for all cameras. First up the Canon on centre-weighted. With the focus and weighting on the building, but taking into account the light and dark zones everywhere else, it was expected to be a little lighter with loss of detail in the sky and that’s the case. With the zone metering system the camera takes into consideration all the zones. Now, the sky zones are much brighter, but there are a lot more darker cathedral ones. In the case of the Canon’s Evaluative metering they have balanced each other out and the result is virtually identical. For the Nikon, the centre-weighted version was quite similar, though slightly darker than the Canon. However, the zone metering version was a lot brighter showing that in this case, the metering has taken more note of the fact that there are a larger number of average, darker ones, than the brightness of the few sky ones. The Pentax has given the darkest centre-weighted image, putting more detail back into the sky and again the zone metering has recognised and counted the majority of the darker zones and not been fooled by the brighter ones. What can be drawn from this is that on zone metering, the Pentax and Nikon will expose for the ground rather than the sky and the Canon takes slightly more notice of the sky. It’s important to realise that there is no ideal exposure here, because the tonal range is too great to be captured in one shot.

The great DSLR shootout backlit, contrasty scene test: Click on the thumbnails for the larger images.
Canon EOS 550D backlighting centre-weighted metering Nikon D90 backlighting centre-weighted metering Pentax K7 backlighting centre-weighted metering
Canon EOS 550D backlighting centre-weighted metering. Nikon D90 backlighting centre-weighted metering. Pentax K-7 backlighting centre-weighted metering.
Canon EOS 550D zone centre-weighted metering Nikon D90 zone centre-weighted metering Pentax K7 zone centre-weighted metering
Canon EOS 550D zone metering. Nikon D90 zone metering. Pentax K-7 zone metering.

Test 6: Bright sky, dark foreground
This is a similar test, where the church front is in shadow but there is a nice swathe of blue sky to the left, with the sunny sky off to the right. So, there should be a good chance of getting both building and left-side sky together, at the expense of the white clouds to the right. This test was done with zone metering and gave some interesting results. The first is that the Nikon has lost the sky to the right while having the darkest building front. The white balance has also given the church front a very cold feel. The white balance of the Pentax and Canon are better. The Canon has brought in more sky, so has a lighter building. However, the Pentax has a darker building than the Canon and yet also more sky, giving a slightly better dynamic range result. So, again the Pentax and Nikon produced exposures slightly more aimed at the building, the Canon took slightly more notice of the sky, but the Pentax gave the best result overall.

The great DSLR shootout bright sky, dark foreground test: Click on the thumbnails for the larger images.
Canon EOS 550D bright sky zone metering Nikon D90 bright sky zone metering Pentax K-7 bright sky zone metering
Canon EOS 550D Nikon D90 Pentax K-7

Test 7: Predominantly black subject
It’s time to do some spot metering here with a predominantly black test. The test subject is an old mill. On one side is blue skies and sunshine, on the other shadow on the black painted surface. Spot metering is designed to place whatever you meter off, in the middle of the exposure so when you look at a histogram, the area that was spot metered from, should be smack in the middle. So, in this case, the test is to see whether the metering ignores everything else that is going on and places the shaded, black surface in the middle of the exposure. The result of that will be a burnt out sky and overexposed picture. In picture terms that isn’t what we want, so the second part of this test is to ask zone metering to balance the bright blue sky, with the black front and very black back of the mill.

First up the Canon and the spot metering has indeed rendered the dark black as a light muddy grey colour. It did fail with a slight blue colour cast but that’s a separate issue. The zone metering also performed very well and captured the entire scene for the Canon. The Nikon spot metering produced virtually the same result so a good performance there and the white balance was better. However, the predilection of the zone metering to pay more attention to the landscape than the sky is clearly demonstrated here. The mill is more of a muddy grey, the sky has lost detail and the image is overexposed. It’s no surprise then to see the Pentax give a similar result. Firstly, the spot metering ups the blacks into the middle as with the others, but not to quite the same degree, showing that there is less weighting there. On the zone test though, the image isn’t that much different, showing more weighting towards the scenery like the Nikon, with overexposure in the sky, though it isn’t as pronounced as the Nikon, showing a better dynamic range. Clear win for the Canon on this test, Pentax second.

The great DSLR shootout predominantly black subject test: Click on the thumbnails for the larger images.
Canon 550D predominantly black spot metering Nikon D90 predominantly black spot metering Pentax K7 predominantly black spot metering
Canon EOS 550D predominantly black spot metering. Nikon D90 predominantly black spot metering. Pentax K-7 predominantly black spot metering.
Canon 550D predominantly black zone metering Nikon D90 predominantly black zone metering Pentax K7 predominantly black zone metering
Canon EOS 550D predominantly black zone metering. Nikon D90 predominantly black zone metering. Pentax K-7 predominantly black zone metering.

Test 8: Predominantly white subject
This dual-shot test is the opposite of the previous one. Here we have a white door in the sunshine. Accurate spot metering should give a dull result with the white door tones in the middle of the histogram. The challenge for the zone metering is to read the duller brickwork surround it and shift the door tones into the top half of the histogram. Firstly the Canon on the spot test and here this is disappointing. The white tones are about a third of the way into the histogram, not in the middle so it’s much too dark. On the zone metering its taken notice of the darker brickwork and now it is brighter, but with almost all the tones under the two third point it’s still underexposed. The Nikon is better, but still, the white door isn’t in the middle of the exposure. The zone metering version is brighter again than the Canon and is a much better result overall. We’ve seen the Pentax tend to give darker results than the other two systems and this is very apparent with the spot metering test on the white door. It’s completely wrong, the door tones are about 25% of the way into the histogram, rather than being in the middle. The zone metering gives a better result but even here, the door tones are now only into the middle of the histogram, so this is very underexposed. Best results here for the Nikon, Canon second.

The great DSLR shootout predominantly white subject test: Click on the thumbnails for the larger images.
Canon 550D predominantly white spot metering Nikon D90 predominantly white spot metering Pentax K7 predominantly white spot metering
Canon EOS 550D predominantly white spot metering. Nikon D90 predominantly white spot metering. Pentax K-7 predominantly white spot metering.
Canon 550D predominantly white zone metering Nikon D90 predominantly white zone metering Pentax K7 predominantly white zone metering
Canon EOS 550D predominantly white zone metering. Nikon D90 predominantly white zone metering. Pentax K-7 predominantly white zone metering.

Canon EOS 550D, Nikon D90, Pentax K-7: Focusing and drive speed

Test 9: Continuous focus and fast shooting
If you want to pick out the killer shot when something is speeding past your lens, you’re going to need the ability to either continually focus or shoot lots of images very quickly, or in fact, both at once. The test here then is to see how well the cameras focus on a car travelling in front of the camera. The test was timed to also include the number of images that the cameras could capture in the same time. The cars, by the way, are entering a town, just past a speed camera sign, so tend to slow down to the same speed. Of the three, only the Pentax has continuous focus marked on the lens as a switch, the Canon uses a joypad button then a menu selection, the Nikon uses a button and wheel combination.

First, let’s just say that continuous focus on all three is tinged with wishful thinking. Something like the cars moving on the same plane across the camera are okay, something coming towards the camera or flapping about like a bird would be impossible. The focussing just isn’t fast enough on any of them. The Canon did have a slight edge in focussing ahead of the Nikon though. In terms of drive and saving to the card, all three have a continuous high speed mode, but here the buffer size also comes into play. The Canon managed a poor six CR2 files before filling the buffer, the Nikon bagged nine in fast succession and then a final one once the buffer cleared to make 10 NEF files.

The Pentax shot more slowly than the others but only managed five PEF files. As a second element to the test, the cameras were then set to top quality JPEG and a 10 second capture test run. The Pentax now managed seven images, still poor, the Canon a far more exciting 18 shots and the Nikon equal with 18 shots as well. So, for action and drive, Canon and Nikon on JPEGs, the Nikon if you want it RAW.

The great DSLR shootout continuous focus and fast shooting test.
Canon EOS 550D continuous focus and fast shooting test
Canon EOS 550D
Nikon D90 continuous focus and fast shooting test
Nikon D90
Pentax K-7 shootout continuous focus and fast shooting test
Pentax K-7

Canon EOS 550D, Nikon D90, Pentax K-7: White-balance tests

Test 10: AWB in daylight
Testing AWB in daylight is interesting because the actual colour temperature recorded can vary significantly, depending on exactly how much sunlight, sun itself or shade is in the scene. So, a number of images shot during different conditions were assessed for all the cameras and these were the results:

Test Canon EOS 550D Nikon D90 Pentax K-7
Sunlight, clouds & shade 4850K 4450K 4450K
Sun in frame 5850K 5400K 4950K
Sunlight and a little shade 5400K 5600K 5000K
Direct sunlight, no sky 5200K 5000K 4550K

So, as far as sunlight goes, it’s rated at between 5000K and 6000K depending on the exact conditions. Typical temperature at mid-day in summer is rated at 5500K. The middle two tests are the ones with the most sunshine in them. The Canon gives warmer results on the whole than the Nikon D90, the Nikon was accurate on the two middle tests, but when shade or lack of sky is a factor, it drops right down. Pentax gave the coldest results of all, coming in well under the other two, even on the two tests with copious amounts of sunshine. The plus point is that you’ll get much bluer skies from this. The point to remember is that if the AWB used by the camera is lower than the actual conditions it will make the image colder. If the AWB used is higher than the actual conditions it will make the image warmer. Overall, the best results were from the Canon and Nikon, but if there’s lots of shade involved, then it’s the Canon.

Just to make life interesting for our cameras there’s also another test here. We have a typical sunny scene, but there are deep shadows under the trees so maybe you don’t trust the AWB because it could be influenced by the shade. So, the white balance was set to the manual Daylight option. Now you can see what the tendency of the cameras is because they use different points for their manual settings. The Canon used a temperature of 5200K, the Nikon 4950K and the Pentax 4850K. The actual results for this image show the Pentax and Nikon being very similar, with probably too much blue. The skies are very nice, but the trees on the Canon shot are definitely more green. If you shoot RAW it’s all a moot point because you can set up the conversion to warm up or cool down shots to your personal preferences. What is shown here is what you get direct from the camera and in these cases my preference would be for the Canon ones.

The great DSLR shootout AWB in daylight test: Click on the thumbnails for the larger images.
Canon EOS 550D AWB in daylight test Nikon D90 AWB in daylight test Pentax K7 AWB in daylight test
Canon EOS 550D Nikon D90 Pentax K-7

Test 11: AWB in cloudy conditions
On to a shot of a house under bright, white cloudy skies. The Canon gave a result of 4950K, the Nikon came in with 4450K and the Pentax with 4600K. The Nikon result is definitely too blue and needs warming up – never mind in the greenery, the front of the house has a blue cast. The Pentax is better, the trees are greener and the red climbing vine is redder, but it’s still too low and too blue. The Canon result is also too cold, but it’s the best out of this lot by a long way.

The great DSLR shootout AWB in cloudy conditions test: Click on the thumbnails for the larger images.
Canon EOS 550D AWB in cloudy conditions test Nikon D90 AWB in cloudy conditions test Pentax K7 AWB in cloudy conditions test
Canon EOS 550D Nikon D90 Pentax K-7

Test 12: AWB under tungsten lighting
In this test of the AWB we’ve got low light, no natural daylight, just some high power tungsten bulbs lighting the display. The Canon gets us off and running with a temperature of 3250K, the Nikon registered 3000K and the Pentax weighed in with 2700K. The Pentax is clearly under the mark here, so there’s a cyan colour cast. The Nikon is the closest, but is slightly under so there’s a slight blue tinge while the Canon is a bit above so it looks warm. Given the material you’d want a neutral or warm image, but in terms of accuracy, the Nikon was closest.

The great DSLR shootout AWB under tungsten lighting test: Click on the thumbnails for the larger images.
Canon EOS 550D AWB under tungsten lighting test Nikon D90 AWB under tungsten lighting test Pentax K7 AWB under tungsten lighting test
Canon EOS 550D Nikon D90 Pentax K-7

Test 13: AWB under mixed fluorescent lighting
The lighting here is a large fluorescent unit overhead with the slightest amount of daylight coming in and hitting the bottom edge of the plinth. The plinth is white, the walls are an off-white colour. The Nikon came up with a temperature of 3700K, the Canon 3800K and the Pentax 3900K. For this test only, the JPEGs you see on the web page have had the levels equalised to exactly the same degree in each case. All three cameras had underexposed this scene, but by brightening them, it makes it easier to see the colour differences. The Nikon is easily the best result and despite being sandwiched between the other two, the Canon has more of a green tinge than the others. So, the Pentax produces a better result than it, even with a higher temperature.

The great DSLR shootout AWB under mixed fluorescent lighting test: Click on the thumbnails for the larger images.
Canon EOS 550D AWB under mixed fluorescent lighting test Nikon D90 AWB under mixed fluorescent lighting test Pentax K7 AWB under mixed fluorescent lighting test
Canon EOS 550D Nikon D90 Pentax K-7

Test 14: AWB using flash
Camera flash is supposed to be balanced at the same colour temperature as daylight, which is around 5500K. However, when using fill flash the camera pays attention to the lighting in the scene as well. So, this scene was actually very dark, with just a little light at the front. The cameras were all set to the same aperture and speed and used in manual mode so that ambient light had little effect. Firstly, the Canon recorded a temperature of 5700K, the Nikon 5250K and the Pentax 4800K. In some ways this mirrors what we’ve seen all along. The Canon clearly is aimed at giving warmer results and has the most red in it. However, all three results are actually acceptable, though the Pentax is slightly on the blue side.

The great DSLR shootout AWB using flash test: Click on the thumbnails for the larger images.
Canon EOS 550D AWB using flash test Nikon D90 AWB using flash test Pentax K7 AWB using flash test
Canon EOS 550D Nikon D90 Pentax K-7

Test 15: Custom WB
For the custom test there’s bright daylight outside the target area and deep shade under it. The test here is to meter off a white sheet of paper to produce a custom white balance for the scene. The Pentax produced a temperature of 4350K, the Nikon 4950K and the Canon, unsurprisingly, 5200K. The Nikon and Pentax are the most accurate in terms of the light on the bricks at the back of the machinery. On the Canon, the brickwork is virtually yellow. On the Pentax the surface of the large wheel looks a lot less rusty than it should be. Of the three, the Nikon is certainly the best result, the Pentax, despite shooting well under what the Nikon produced, is the next best result, and the Canon, with its tendency to warm things up, is too ruddy, but all three are fairly good outcomes.

The great DSLR shootout Custom WB test: Click on the thumbnails for the larger images.
Canon EOS 550D Custom WB test Nikon D90 Custom WB test Pentax K7 Custom WB test
Canon EOS 550D Nikon D90 Pentax K-7

Canon EOS 550D,
Nikon D90, Pentax K-7: Lens quality
For this set of tests a scene was shot at the wide angle 18mm, around the middle 35mm and at the telephoto 55mm settings on the kit lenses. With focal length multipliers of 1.5x or 1.6x, the actual field of view for each was 27-29mm, 52-56mm and 82-88mm. At each focal length a full range of aperture shots were taken, so you can compare any of the settings with each other. As perceived sharpness is affected by brightness and contrast, when looking at this all the exposures were adjusted so that they were the same. All the cameras were mounted on a tripod for this test, though the ISO wasn’t increased in order to see how well the image stabilisation for each system also did.

Test 16: 18mm focal length
At the wide angle with the open aperture of f/3.5 there’s much the same barrel distortion on all three cameras and all three feature colour fringing in equal amounts. One noticeable difference – other than the overexposure from the Nikon, is that the sky has better smoothness from the Pentax than the Canon. At f/8, the sharpest part of the lens, the Pentax just edges it, the other two are inseparable. At the f/22 aperture there is again, no appreciable difference in the sharpness though the Pentax has a little more contrast.

The great DSLR shootout lens quality - 18mm focal length: Click on the thumbnails for the larger images.
  Canon EOS 550D Nikon D90 Pentax K-7


f/3.5


f/4


f/5.6


f/8


f/11


f/16


f/22

Test 17: 35mm focal length
Moving in to the middle of the lens around the 35mm marking on the barrel and there’s a little distortion from the edges, but not a lot now, on images from all three cameras. At the f/5.6 aperture the Canon lens is slightly sharper than the Pentax and the Nikon which are the same, though the Pentax again has more contrast. There’s possibly slight movement on the Pentax for this shot because, as you move into the middle of the aperture settings, the Pentax clearly retakes the lead and also, has more sharpness at the edges of the image than the other two.

The great DSLR shootout lens quality 35mm focal length: Click on the thumbnails for the larger images.
  Canon EOS 550D Nikon D90 Pentax K-7


f/5.6


f/8


f/11


f/16


f/22

Test 18: 55mm focal length

At the end of the zoom on the kit lenses there’s a little slanting inwards on the Pentax lens but all three now show pincushioning in the middle to a modest degree. At the f/5.6 aperture, the Pentax edges the Canon and then the Nikon in terms of sharpness and contrast. By the time the aperture is narrowed down to f/22 it’s a test to see which system can deliver a decent image at a slow shutter speed. Of these, the Canon was the only real success, there’s definitely movement on the Pentax and Nikon systems. In fact, while that’s not a real surprise here, there’s also some movement at other settings for all three cameras. At 0.3secs shutter speed it can be excused, but at 1/100th sec, with a standard focal length and being tripod mounted, it’s not such a great performance. The image stabilisation for all three with these lenses isn’t impressive. Overall though, the consistent result was that the Pentax lens was sharper than the other two.

The great DSLR shootout Lens quality - 55mm focal length: Click on the thumbnails for the larger images.
  Canon EOS 550D Nikon D90 Pentax K-7


f/5.6


f/8


f/11


f/16


f/22

Canon EOS 550D vs Nikon D90 vs Pentax K-7: The conclusion
When testing over numerous criteria as has been done here, it would be unusual for one camera and kit lens to be best at everything and that’s certainly the case. In terms of pure image quality, at ISO100, there’s no real difference between the cameras. As the ISO is increased the Pentax quickly shows more noise than the others so that by ISO400 it’s much more apparent in the shadows. The noise suppression on all three cameras is to be complimented, with perhaps the edge going to the Canon. The Pentax lens though has the most contrast and sharpness at a basic level. For white balance there’s a clear difference of opinion, the Canon gives you a warmer image than the Nikon or Pentax. In terms of outright accuracy, the Nikon is the best here. It’s also the best in terms of continuous or drive shooting in RAW, and tied with the Canon when using JPEGs. When it came to metering, in a variety of conditions, both inside and out, the Nikon and the Pentax were more reliable, with the Canon proving erratic at times. However, the Pentax clearly showed better dynamic range without enhancements, producing detail in the shadows, yet also coming up with generally more detail in darker skies. The Nikon was the worst in terms of dynamic range, losing highlights that the others captured. This comparison hasn’t covered handling but a quick word here is that the D90 is very good with a wide range of buttons and solid build quality, the Pentax is next best being surprisingly heavy, but relying on menus for some things and the Canon is some way behind, with a plastic feel and lots of menu access.

So, is there an overall winner from the test? I’m afraid not. The Pentax is as good as the Nikon and Canon at ISO100, and better at a number of things, but the image quality drops off when the ISO goes higher. If you are expecting to routinely shoot at high ISOs, then either the Canon or Nikon are good, the Canon also has an extra high mode that only really suits black and white. For pure landscape shots or if you just want to shoot JPEGs in sunny conditions I’d go for the Pentax, while for people shots and landscapes featuring stonework you might want the warmer tones of the Canon. For more action or mixed lighting conditions, such as indoors, then the Nikon would be my choice.

The Canon EOS 550D with 18-55mm lens costs around £643 and is available from Warehouse Express here:

Canon EOS 550D

The Nikon D90 costs around £609 (the 18-55mm lens costs £147) and is available from Warehouse Express here:

Nikon D90

The Pentax K-7 with 18-55mm lens costs around £788 and is available from Warehouse Express here:

Pentax K-7

Explore More

Join ePHOTOzine and remove these ads.

Comments


Bonico 4 Romania
23 Aug 2010 1:01PM
Very interesting review, I would love to see more of these!
Bogdan

Join ePHOTOzine for free and remove these adverts.

YoBellzaa e2
7 224 England
23 Aug 2010 1:02PM
Excellent, we plan on doing more Smile
timbo 12 591 United Kingdom
23 Aug 2010 2:42PM
for me this is no real surprise. The D90 is a jewel among the Nikon range at a fraction of the price of 'higher' spec models
konu 6 31 United Kingdom
23 Aug 2010 5:54PM
The result was not surprising and I like the test, waiting for more like this Smile
Kako 8 140
23 Aug 2010 7:37PM
Excellent review, the only proper way you can compare group contenders-at the same time under near identical conditions. If I was looking to choose from these contenders this is exactly the sort of side by side test I would place great store in. Ultimately I would still want to handle the cameras for myself to see how they felt in my own hands, but at least I would be well informed on the technical aspects of their performance and thus better able to come to an informed decision.

If I had a criticism (only a small one!), could we please have these comparative detailed tests soon after the cameras come out -not when they are on their way out or have been around for many months.

Overall, great job.
HouseMartin 5 193 1 England
23 Aug 2010 7:45PM
Whilst I would generally agree with these test results (there is no denying the D90 is a cracking camera body) I was a bit worried to see the glaring technical error in that the Pentax does actually start at ISO100, not, as highlighted, 200. One error questions the possiblity of others. Also, if you used NEF on the Nikon why did you not use PEF on the Pentax? The feel throughout this was an unfamiliarity with the Pentax and Pentax range in the hand, and a knowing nod to generally good Canons.

Finally, would have been nice to have seen the Canon and Pentax compared to a D90 with an 18-55 kit lens attached, not a VR (albeit budget). Just about everyone upgrades the kit lens soon after buying.

This site is generally justifiably trusted in its reviews and opinions. A pity this one was not so rigorously performed. A nice bit of Nikon Nation PR possibly? from what could have been a genuinely unbiased review, with research, showing the good points of all and in conclusion the D90 being the generally better camera.

Paul
PS I do own a Pentax (not a k7, prefer the K-x) but my comments would have applied equally for the other two if they had been mis-described. Also totally agree with kako, bottom line - go play with them before you commit.
Another correction re the Pentax K7. Like the Nikon D90, it also has user configurable high ISO noise reduction with the same four settings of off, low, medium and high.
brian1208 e2
11 10.2k 12 United Kingdom
24 Aug 2010 7:42PM
I notice that the canon 550D comes in at around £100 less than its two competitors a fact which doesn't seem to be mentioned anywhere in the review. Spend that extra money on a better lens than the kit lens and I wonder what the result would be?
master43 10 4 United Kingdom
24 Aug 2010 10:03PM
Then explain how a respected camera magazine rated the Pentax K7 ABOVE all 3 for best overall DSLR in a recent test?? Clearly there is an element of subjectivity in this test! The images of ALL three cameras are just as good as ANY other DSLR camera, and all will produce high quility images good enough for publication!! The only point I tend to agree with is that noise control at high ISOs in the Pentax is not as good! But, as stated in the previous comment, the noise reduction is customizable. And of course the lowest ISO for the Pentax IS 100 ISO!!

What about the build quality of the Pentax K7? It's build quality is far superior to the other models! And handling?? Well, unless you have hugh hands it is also more ergonomic than the Canon and Nikon.

I'm not surprised at the rather cursory glance and instant dismissal of Pentax as this site is biased towards Nikon and Canon anyway!
Pete e2
13 18.4k 96 England
24 Aug 2010 11:25PM

Quote: I'm not surprised at the rather cursory glance and instant dismissal of Pentax as this site is biased towards Nikon and Canon anyway!

Utter rubbish! I use Pentax and run the site.
25 Aug 2010 1:13AM

Quote: Quote:I'm not surprised at the rather cursory glance and instant dismissal of Pentax as this site is biased towards Nikon and Canon anyway!Utter rubbish! I use Pentax and run the site.

I recently traded a K7 and bought a D700 with a D90 for backup. I like the D90 and I agree with your conclusions regarding high ISO noise and continuous AF.

However the Pentax does have a lot more features (not mentioned) and far better build quality (metal body and sealed). Handling on the K7 is really superb with far better access to custom settings, esp WB tuning and auto ISO settings. Again, not mentioned. In many ways its better than the D700 (no setting for auto ISO to adjust itself based on focal length for instance).

As far as I can remember it also does ISO100 perfectly well and has four NR settings for high ISO. These innaccuracies are surprising. The Nikon OTOH only does ISO100 as a non-standard setting. Auto WB is also a considerable improvement over the Nikon in incandescent light.

The Nikon has many advantages in terms of continuous AF and is easier to live with overall, but you dont need to put the Pentax down for faults it doesn't actually have. I won't accuse you of bias, but at least get your facts in order.
25 Aug 2010 1:28AM
The review was a little disappointing in that there's a bit of difference in the kit lenses.

As each of the contenders have have a decent/affordable 50mm prime that almost everyone buys, it would be nice if you did the tests with that or a similar lens.

As it is when I see the conclusion state that the IQ is really close but the D90 is better, and then look at the differences in lenses, I have to wonder if it is the body or the glass, especially given the 18mp sensor on the 550D/T2i does not really get a chance to shine with the kit lens....
25 Aug 2010 1:37AM
This review is a mess with regards to the K-7.

There are basic factual errors: The K-7 shoots ISO 100, if you can't access ISO 100 you most likely have shadow and highlight correction turned on. This a a software setting for JPEGs which will affect images considerably with regards to dynamic range, noise and overall exposure. You may also have had things like CA correction and lens distortion correction turned on? Again a JPEG software setting which will change the ouput, buffer etc.

And yet you haven't mentioned these settings. Like you don't mention in-camera shake reduction, weather sealing (camera or lens), 100% viewfinder, older lens compatibility... the list could go on.

Are you normally shooting with all the settings at default in-camera JPEG? It's pretty much not testing how noisy the sensor is here, but the default in-camera noise reduction. Proper tests show the K-7 in RAW is very similar to D300S and 7D, with perhaps slightly more noise and detail retention.
And the Nikon had quite a different lens.

And I have no idea how you managed FOUR shots in continuous shooting before a buffer jam. My K-7 with a 'fast SDHC' card shoots about FIFTEEN RAW files at 1/8000 second (so with no frame-rate slowness due to shutter speed) before any buffer issue. No indication of number of in-focus shots on here either.

No attempt to discuss any other factors like build quality, audible noise, ergonomics etc etc.

A very poor review. Perhaps not biased, but factually a mess.
My advice would be do a review properly, or don't review.
leeak 4
25 Aug 2010 2:58AM
When you post a messy review as such, your site loses credibility and viewership. Count as one less visit from me in the future ....
ramiot 4
25 Aug 2010 3:40AM
I am in full agreement with the forum members that are critical of your review. How could you have not noticed that the K7 has a ISO 100 native. Your fast shooting analysis seems realy out of wack. Your review is not a serious analysis and it would be better not to publish this poor quality analysis if you want to keep your credibility...

Robert Amiot
Québec Canada
25 Aug 2010 5:46AM
This is one of the least credible reviews I've ever read, and it certainly draws into question the credibility of the other reviews on the site. Did you actually have a K-7 when you conducted these tests? Good grief.
Thanks for posting. I'm sure the efforts put into producing tests like this and then reading comments here must often seem a thankless job.

I think some of the less forgiveable aspects of this "test" is that you shot nearly all of the K-7 images using center-weighted metering rather than segmented which would have been the equivalent of Evaluative/Matrix on the Canon/Nikon. (No wonder they were so 'consistent'.) This led to significant underexposure on many images.

The other is the apparent lack of effort on the part of the tester to familiarize himself enough with the Pentax to know that ISO 100 is not accessible if the D-Range/shadow correction features are enabled...another feature that will sometimes introduce some noise into the shadows.

So combining use of shadow correction and underexposing by misuse of center-weighted metering puts the K-7 at a significant disadvantage that it didn't need against formidable competition.

I understand why the D90 was tested with its pricier 18-105 kit lens rather than the 18-55 which would have made for a better comparison...don't think this would make that big a difference anyway. Certainly didn't help in terms of distortion.

As a suggestion for future tests of this sort on lenses, at the vary least you might want to include some relevant 100% crops that made you decide which lenses were 'sharper'. It can be a bit tough to judge these anyway as the scene had objects at many different distances so focus may not have been the same.

The 'write error' on the K-7 during continous shooting doesn't sound like normal behavior to me and is probably an issue that should have been resolved rather than assuming 'that's the way it is'.
25 Aug 2010 6:09AM
All I can say is I am extremely disappointed that something this poorly done could make it to the presses on a major photography site.

If in fact Pete, the site owner, does own a Pentax than he would look at this head to head and scratch his own head.

Garbage like this damages this sites credibilty. When I initially clicked the link, which was posted on photo.net Pentax forum, I saw ephotozine and I expected a quality comparison, much in the same way I'd expect a good review from photo.net, dpreview, etc. From this point forward this site NEVER gets the benefit of that again.

So Pete, if you do have any control over what is posted on your site, you should be considering that I am most likely not the only person that feels your site has lost most of it's credibility in one quick instance. I would highly recommend you pull this review until the author can spend some quality time with all 3 cameras.

2 major inaccuracies, which just a quick glance at the spec sheet would alert the author to, are the RAW buffer, and the base ISO. Just these two major errors alone are enough to crush the credibility of this review.

Initially, I thought it was a typo that the K-7 didn't have ISO 100, but when I scrolled to the first set of test photos I noticed the reviewer really didn't test the K-7 at ISO 100.

Beyond that, the K-7 metering is quite good, yet the reviewer managed to underexpose almost all the K-7 images by what appears to be close to a stop. I wonder if along with DR expansion, he also had exposure compensation set on the k-7 to -2/3 or -1.

With any speed card the K-7 buffer is 15 PEF or 14 DNG, using the fastest cards I have I actually get an extra few shots off before the buffer fills, as I always have with Pentax DSLRs since the K10D. This makes the buffer more like 17 RAW vs. the 4 that the review states. This isn't a minor discrepancy, it's absolutely appalling that the reviewer didn't notice this and try to figure out why he might be getting such limited buffer from a camera with quite a large RAW buffer at 5.2 FPS.

These are the errors (the buffer and ISO) that rule out subjective influences because they are simply specifications of the camera and not open to debate. There are other areas of the review that I question, such as the metering where the reviewer notes Pentax was the most consistent, but Nikon wins? Huh? How is this possible?

The bottom line, you've lost a few people who might have looked at your site as a source of information, and I truly hope you can correct these issues before your site loses more credibility. I hope you realize that even one article on a website as poor as this can call into question everything else on the site.

If your reviewer needs a crash course on the workings of the K-7 before reattempting a less biased test, he can feel free to take a look at the thorough review of the Pentax K-7 found on photo.net
bkpix 4
25 Aug 2010 7:11AM
Very sloppy work on the reviewer's part.

Given all the obvious errors, I suggest you pull the review until someone can update it with correct information.
Spag 4
25 Aug 2010 9:42AM
I can only agree the above posters, very poor job from the tester. Never thought I'd read such a bad review here. Sad
Pete e2
13 18.4k 96 England
25 Aug 2010 10:23AM
Thank you for all joining to share your views. There's clearly some Pentax bias in some of your opinions, but this also highlights clearly that the review is flawed. As you're all new to the site you won't be aware that we are normally very accurate with reviews and this is a one off which we will address. We pride ourselves with not having any bias which is why I stepped in earlier to confirm this. We will re-evaluate the review and reshoot some of the test pictures making any necessary changes to the review.
Pete e2
13 18.4k 96 England
12 Nov 2010 2:19PM
It's taken a little longer than anticipated but the above is now a total rewrite based on a more thorough understanding of the cameras.
JackAllTog e2
5 3.6k 58 United Kingdom
16 Nov 2010 5:43PM
Excellent review & summary.
If this was to follow with the 60D, D7000, K5, Alpha55 i'd be chuffed to bits.
rob27 4 2 United States
5 Oct 2011 8:55AM
A really interesting test design, but dissapointed that most all subjects are inanimate stuff like buildings, bridges, streets, etc. At least for me, the test of any camera is how it handles the most challenging and interesting subject matter of all...people. Particularly faces, skin tones, etc.

Scenery is nice, but people are hard-wired to respond at the deepest level to the human image. For the next test, maybe consider showing us what a camera can do with human subjects.

Once again, the study design is good, but the range of subjects is a bit too clinical, impersonal, technical and less than engaging. - Rob

Sign In

You must be a member to leave a comment.

ePHOTOzine, the web's friendliest photography community.

Join For Free

Upload photos, chat with photographers, win prizes and much more.