Images from the Nikon D800E

25/04/2012
Okay, so I’ve postulated over the past few blogs about the handling of the Nikon D800E. I’ve wittered on about what I see with the camera in the field, but all that is just about the camera in the main, not about the images. So that’s where we are now. Having taken the images, what are they like?

The opening image has received no processing, no sharpening, (not even in ACR), so is just a coversion to JPEG and downsized.

First, to state the case. The images were all taken in 14bit RAW as opposed to JPEG. (12 bit can be chosen if you prefer.) The initial images, which I will talk about here, were all taken at ISO 100, the camera’s native ISO and therefore at its best. The camera was set on a Gitzo tripod with Manfrotto 410 Junior geared head with both mirror lock up and a remote release used. All the lenses used were Nikon lenses. I deliberately left my Lee Filters in the rucksack to really test the dynamic range. DXOMark gave it a stonking DR of over 14 EV - that’s getting close to us humans visual ability when in our prime, if certain proclaimer’s are correct.

Anyway, first things first, opening the images. Now, since it first came out, I have used Adobe Lightroom as my RAW converter, catalogue and filing database. Problem number one. Lightroom 3.6 does not support the D800E. The new Lightroom 4 does, but Adobe say that it won’t work in all circumstances at the moment, something to do with the live view. Fortunately there is the free download of Nikon ViewNX which can be used and is available to all, plus, with purchase of the D800E, there is the Nikon Capture NX2 software bundled free via a licence key. (Post note: I have since downloaded Adobe ACR 6.7 for my CS5, but it is only a Beta version and not tried by me yet.)

I downloaded both CaptureNX and viewNX and used the transfer engine in viewNX to import the files which I then opened them in CaptureNX. As with most RAW converters, luckily they all follow a similar format. CaptureNX is probably a little different, because of its colour control points, but it didn’t take long for me to see the images and do the basics. I deliberately wanted to keep the files pretty much free from all but necessary processing. In fact, the opening image has no processing other than file conversion from RAW to JPEG.

When the first image opened, it was immediately obvious that it had detail, and loads of it.. Then I viewed it at 100% on the screen. Awesome. More detail than I have ever seen before in an image, as expected I suppose but nevertheless gobsmacking. Scrolling from top to bottom showed that the image wasn’t perfect by any means. To be fair, my first takes hardly ever are, they all need some correction factor in the field, either focus position or exposure compensation.

Looking at the histogram, even without any neutral density graduated filter, all the detail in the sky was retained. I will go on to cover the DR later, but suffice to say it is impressive. The little D7000 always amazed me in this regard, far better than the pro body, which I know is old hat now, but nevertheless, impressive.

It is always a bit of a danger getting bogged down with images at 100% on the computer and comparisons with other cameras performance. Why? Because the more megapixels the camera has, the larger the image you are viewing and therefore you are not comparing like for like in this respect. That’s why DXOMark standardises the images in its tests. Moving on.

A number of things became blatantly obvious while looking at these first images. One, they are bloody HUGE! Nikon quote uncompressed RAW files at 74.4mb each. Now, my computer is running the ram saving XPpro, with 4mb on its board, controlled by a 2.4gb dual core processor. Even with this set up, I didn’t have more than two full screen files open at a time, that’s for sure. Saving one of these mamas as a tiff file took circa 45 seconds on my machine, and those Tiff files? They come out at a jaw dropping 207mb EACH in 16 bit. Now, I was expecting something like this, but the purpose of mentioning it is to alert you, should you have a lesser computer with minimal hard drive space left, think about an upgrade or new computer!

The second most salient point is because you can see so much detail, that not only includes every blade of grass and every twig on a tree hundreds of yards away, but also every flaw in your technique, every slight frequency change of your tripod legs caused by the wind, every idiosyncrasy of your chosen lens.

To highlight the point, the D800E has 7360 pixel capture capability. Viewed on the screen at 72 ppi, what you see, (if you had a monitor large enough), would be an image 8 foot 6 inches wide! Hence you have to be realistic about what you see and you don’t view an eight foot wide image from 15 inches away. (Yes I know an image like this wouldn’t be printed at 72ppi but you get my drift hopefully.

On location I took a number of different images which, using a couple of different lenses, I shot at different apertures so that now on the computer I can do some direct comparisons on small sections of three images at a time, viewed side by side at 100%.

The conclusions are not earth shattering and something us photographers all know deep down. That is, your lenses work best between F8 and F11. Let’s take some specifics. With the 24-70 F2.8 lens, both at 24mm and 45mm, at F8, the image was the sharpest without question. But, only within its depth of field at this F8 aperture obviously. Fall off acutance outside the DOF, visually assessed by me, became apparent, but I was happier with this aspect viewed across the entire range than I was when I reviewed the image at F16, and at F22 don’t even bother!

At F11, the depth of field was very acceptable across the image. It was not as sharp as at F8 in the sweet spot, but pretty good. (I only stuck with whole apertures as I had a lot to get in by the way.) Now, to make sense of this, a quick word about what is ‘acceptably sharp’. We are all different. Camera manufacturer’s DOF scales work on the distance covered by the scales being acceptably sharp. But acceptably sharp for the criteria they use is not acceptably sharp for many. As many of you know, this aspect, circles of confusion and hyperfocal distance are pet subjects of mine due to the bollocks spoken about it – all theory and no practice in the field where it matters. I can only tell you what is acceptably sharp to me. My eyes were checked last week and they are at least 20-20 vision, I asked. So it is on that basis that I make these remarks. You will have to make your own judgement.

Back to the testing. Using a 24mm prime Nikon AIS lens, the same numbers came up as best. One of this lens’s characteristics came out though, the softening of the image at the edges.

Now, I’m not saying that these are new problems. I am not saying that the lenses are not up to the task, it’s just that because the sensor resolves so much detail and, more than that, we can see this detail clear as day, you have to be pragmatic as to what you are seeing given the comments I have made above.

Okay, I’ve included some images and crops for you to peruse. Suffice to say, the camera produces masses of detail. I’ve subsequently printed a number and the level of detail, the level of inherent sharpness in the images without any added in processing is beautiful. These prints have a feel reminiscent of medium format without a doubt, a certain clarity that I have not experienced with digital before. However, shoot with care or you will waste your time. Well, in fact, that’s not true, because if you downsize your images to display or print, then they will be more than fine because the ‘errors’ that you see at full size will not show, which is why you haven’t seen them in your images before!

To come in future blogs: How the D800E performs regarding dynamic range and also with noise.


You will probably have noted the crop marks in the opening image, well here are the images produced as crops from that original image. If this enlargement was spread across the whole image, it would be over 8 feet wide!
As I said above, these first two images have no sharpening applied whatsoever.

DSC 0091crop


DSC 0091-2


...and with minimal sharpening
DSC 0091-2 sharpened



Keep practising
I am