When choosing a camera we’re usually confronted with a few choices, Brand and model usually are the main 2. Nikon have added another sub-choice. The D800 or D800E. A few manufacturers have offered sub-choices before, offering a very specific version of a model for a very specific task, the Fujifilm FinePix S3 Pro UVIR for example. A version of the S3 Pro camera from 2006 specifically designed to make photos in the Ultraviolet and Infrared light spectrums.
Nikon’s decision to make 2 variations of the D800 isn’t quite as dramatic but nevertheless it has caused many to think more than twice before committing to buy.
On paper the standard and E version differ in one small but significant factor. Simply put, Nikon haven’t fitted an anti-alias filter into the D800E that deliberately softens the image that the sensor sees. Sensors with anti-alias filters have become the norm in DSLRs and we don’t usually even think about them. You could ask why there is a filter there in the first place, especially when we crave the sharpest images possible. The answer is simple. The filter is a compromise between image sharpness and the risk of the pure unfiltered sensor creating artefacts, or ‘Moiré’ in the final image.
Armed with a D800 and D800E with identical settings and lenses, I set out on my own personal quest to see if I could justify to myself the +/-400 Euro difference between the Nikon D800 and D800E.
I’ve been using the D800 for wide variety of projects and never noticed any Moiré in the images and to be honest, never really thought that the images were not as sharp as they could be. On the other hand, I have never looked for Moiré or assumed the images were not sharp.
For this testing I had to deal with my pixel-peeping phobia and really look for the Moiré phenomenon and to see if there was a noticeable difference in sharpness between the 2 cameras.
For my first outing I went to the city to do some cityscapes, random locations and views to try and trigger any differences that the cameras would reveal. Regularly patterned brickwork or roof tiles are a good way to try catching any issues.
Both cameras were mounted next to each other, settings the same, same lens, same focal point and both cameras triggered at the same time. Comparing the images on the LCDs, there was nothing in it. Impossible to see any difference.
Only under intense scrutiny in Lightroom, with the D800 and D800E files side by side did I detect some Moiré pattern in the D800E files. Some very distant details of roof tiles were showing the typical coloured patterns, but that was all. The D800 files were clean. The same was true for just about all the shot. At 100% zoom the D800E files sometimes had some Moiré creeping in. Only at 200% plus was it really getting obvious.
Then I stumbled upon one shot that happened to have 2 people in the centre of the frame. One was wearing a striped shirt. At 100% the D800E image showed some serious blocking and confusion at the details on the shirt. I put the D800 file next to the D800E file and it to my surprise, it too was seriously affected. I had to double-check to make sure I hadn‘t selected another D800E file but no, it was the D800, very aggressive and disturbing artefacts.
For the 2nd test I enlisted the help of Stijn to pose for a few shots in just about the most challenging patterned clothes we could imagine. Immediately Stijn and his Moiré magnet trousers were sending the D800E crazy. The distinct patterns in the fabric even caused a moiré effect to be seen in the LCD. Every shot of Stijn provoked artifacts in both cameras. The only time he didn’t was when he was in the distance; obviously far away that even the D800’s couldn’t resolve the detail of his trouser pattern.Back on the PC, both cameras were suffering from the discolouration and artefacts. At 1:1 zoom the D800E was noticeably worse. At 200 or 300% both had it bad but the ‘E‘ had me really worried. Worried how I would deal with this in a real life shoot.
Luckily, Lightroom and Nikon’s own CaptureNX2 come with a Moiré filter that can be locally applied and both tools work a wonder on the affected areas. Mostly.. In the case of the cityscape image with the striped shirt, the actual pixel pattern was disrupted so much that some additional post processing would be needed if the D800E photo was to be used a full resolution on a large print. Having 36mp files is a blessing and a burden, the detail shows up any flaws very quickly. If the image was to be output at about 5000pix wide, a lot of that moiré pattern and problem vanishes. However, we have a 36mp camera for a reason. To use it when we need it.
This was the reason Nikon left the anti-Alias filter out of the D800E. To squeeze the maximum sharpness out the sensor. Honestly, in the real world I found it very hard to see any difference at all. I’m sure on chart tests there is a difference. There must be. Mustn’t there? However you really have to see the images side by side to actually believe the D800 is any less sharp than the D800E. There are a couple of times I have spotted a difference, usually on distant text in images or a hard defined edge of a building. For most work though, it’s nice to know it’s there but it’s not going to be noticed if it’s not.
The deep pixel level difference can be seen during processing though. Zoom into levels of 300% and more and the D800E’s pixels are very defined and hard compared to the softer more fluent D800 pixels.
I’ve also noticed a consistent difference in white balance between the 2 cameras. Despite the settings being identical in Lightroom ,the D800E images look slightly warmer in tones. Probably another effect caused by the lack of anti-alias filter. No big deal but could be important when processing D800 and D800E images together.
What to choose?
This is a question I’m bombarded with and I have battled with the same question.
It has been made clear to me that both cameras are susceptible to Moiré, probably due to the sheer level of detail they record. All but the most aggressive Moiré can be worked out of the image very quickly using any one of many of the processing applications available to us.
I can see 2 buying groups here. The photographer who is very detail minded would quickly opt for the D800E. Serious moiré creeping in might be a shame, but not a disaster.
Then you have the photographer who needs to be sure their shots are good and the processing workflow is efficient. If they’re heavily shooting Macro or fashion, they might opt for the safe low risk D800. Saying that, I know people who haven’t taken the safe route and not had problems with the D800E.
Nikon have given us a choice and warned us in no uncertain terms about the pros and cons of going with the D800E.
I have no hard answer to anyone wondering which to buy. Are you going to miss anything by buying a D800? Not really.
Are you going to gain anything by buying a D800E? Slightly.
Is the risk worth it? Only you can know based on the type of work you do.
Personally, I like to live on the edge. I took the D800E.
(I will ask Stijn to wear different trousers next time though…)
Below some screenshots of my test shoot. Under the screenshots a link to download a .ZIP of the full resolution .JPGs talked about in this article.
At 2:1 zoom Colouration in the roof tiles and bricks is quite prominent in the D800E (left) image and very slight in the D800.
The D800E (right) does definitely have sharper images. The difference is minimal but can clearly be seen in fine details such as this architectural comparison.
This is scary example of how bad things can get. The detail might be minute in the overall image but if this was being used a full resolution, both the D800E (left) and D800 (Right) images would need some extra processing.
Stijn’s trousers really did test the cameras as I wanted. Strong repeating patterns cause both cameras to suffer quite heavily from Moire.
Luckily all but the more aggressive Moire pattern can be quickly removed with moire fix tools found in most RAW processing software.
This shows that Moire isn’t always a huge issue. The same trousers shot at a different angle and although both cameras are displaying Moire, it is a lot less.
Whichever camera we choose, we are spoiled for detail and sharpness levels.
Link to to the high-res images used in this test. (2x zip files ±75mb each)
This article was originally published in Shoot magazine, edition 21. The Belgian photography magazine.
Thank you to Erik for allowing me to reproduce the document in for the blog.
Thanks also to Stijn for giving me a hand with the testing.
For more regular updates on what I’m up to in my job, please feel free to join my page on Facebook.