Nikon has its Mark III

[BernardLanguillier]

Thank you Bernard for your feedback. Your comments are really important for people like me who are still debating whether to buy a(n other) D800/E or the new D810. Sure, we can always read the reports like the DXO one, but the input of "real-life" professionals is invaluable.

It really depends how important an expense the delta of cost is for you in the grand scheme of things.

Honnestly, the D800 remains an amazing camera.

Cheers,
Bernard

[BernardLanguillier]

The 810 has only half a stop more DR than the 800E - and it's less than half over the 800.

The 800E has the best Low ISO score of the three cameras.

Yes. Now look at what Canon did over the last 6 years.

http://www.dxomark.com/Cameras/Compare/Side-by-side/Canon-EOS-6D-versus-Canon-EOS-5D-Mark-III-versus-Canon-EOS-5D-Mark-II___836_795_483

Cheers,
Bernard

[Lupin]

Now look at what Canon did over the last 6 years.

Not as well as Nikon judging by those figures. 

[MrSmith]

but what about these figures? 

[bcooter]

I don't think any of these charts talk about image quality as much as market penetration.

Canon had a full frame camera way before Nikon, way before anyone and if you were a professional 8 years ago, it was Canon or medium format or both.

It took Nikon forever to come out with a full frame camera and when they did, it was aimed more at the sports photographer than the general advertising market.

Consequently a lot of people have a lot of legacy Canon glass.

Also Canon jumped on the low cost video bandwagon early with the 5d2, continued on with the Canon mount on the 1dc, the c100 to 500 and most third party cameras like RED have a Canon mount solution.

Nikon is just beginning to start promoting video and now offers a "kit" with the d810, along with a marketing campaign, which in 2014 is kind of like someone at Nikon said, "you know video might be a big market", which kind of makes you wonder what they were thinking in 2009.

Once again, I'm not judging usability, color, detail, image quality, just market density and Canon had a very long head start.

In fact Nikon had a leg up in cinema and video in the adoption of their lenses.  For decades the only still camera lenses used for cinema and video were basically Nikon and older Nikon lenses like the 50mm 1.2 still go for a decent price.

Why Nikon didn't capitalize on this is confusing, especially since they had no ENG camera market to protect.

So bottom line, of the image makers I know that shoot combination projects, few don't leave the door without some kind of Canon product, whether it be cameras and bodies or tilt shift lenses.

Canon covers a lot of territory and once again this isn't meant to start flaming any other brand because I own both and don't sell cameras for a living.

IMO

BC

[BernardLanguillier]

The only reason why I linked to those Canon data is to highlight the fact that increasing DR by half a stop when Nikon/Sony was already this far ahead of their main competition on that indicator is a pretty impressive technological feat. This was just a reaction to the "only 1/2 stop?... " comment.

It doesn't mean that Canon cameras aren't great tools or that Canon is going bankrupt. It was not meant to be about Canon. They were just meant to be a data reference point.

I do hope as much as anybody else that Canon will come up with a revolutionary sensor technology at the Kina putting Nikon years behind. That would be a perfect base for my collection of lenses. One can of course wonder whether that would have any practical value considering where Nikon is today relative to actual needs. At 14.8 stops DR at base ISO with 36mega very sharp pixels and great colors, most architecture/landscape contrast handling needs are covered very well but more DR at higher ISO would help in many situations such as wildlife, concert photography,...

Cheers,
Bernard

[Ray]

A half stop improvement in DR is considered by DXO to be the minimum improvement that has significance.

I recall when Nikon brought out its first full-frame DSLR, the D3, almost everyone raved about its high-ISO performance, which was claimed to be significantly better than any other DSLR on the market. However, DXOMark test results later showed that the D3 had just a 1/2 stop DR advantage at ISO 1600 and above, compared with the older Canon 5D model with similar pixel count.

If one adds up all the incremental improvement in the D810, such as electronic first curtain which should improve image quality at slow shutter speeds, more accurate autofocussing, faster frame rates, slightly higher DR at base ISO, better implementation of LiveView, and possibly a slightly sharper image due to the complete absence of any AA filter components which otherwise might introduce some degree of image degradation, which is a possibility in the D800E, then the sum total of all the minor improvements could represent a sound reason for buying a D810, especially for someone who has had an eye on the D800E but has resisted buying one.

[HarperPhotos]

Hello,

In the comparison test chart at the link below.

http://www.dpreview.com/previews/nikon-d810/7

When I look at side by side images of the new Nikon D810 and Nikon D800E  in Raw and 100 ISO what I see is the the D800E is equal and in some chases sharper than the Nikon D810.

Just my point of view.

Cheers

Simon

[Simon Garrett]

The image quality differences are fairly slight, and I suspect that the size of some of the differences may well be comparable with sample-to-sample variation and measurement error.  Looking at the dpreview raw test images and the DXOMark tests, the only noticeable difference seems to be better DR at ISO 64, and perhaps slightly lower chroma noise (amp noise?) above ISO 3200. 

But I guess few people will change a D800/e for a D810 for image quality - it would be the other improvements that would attract. 

[Lupin]

Hello,

In the comparison test chart at the link below.

http://www.dpreview.com/previews/nikon-d810/7

When I look at side by side images of the new Nikon D810 and Nikon D800E  in Raw and 100 ISO what I see is the the D800E is equal and in some chases sharper than the Nikon D810.

Just my point of view.

Cheers

Simon

I agree. The D800E 100 ISO looks better than the D810 100 ISO - it's got smoother detail.

If you blow them both up to 200% and look at the centre circle on the chart, the D800E NEF is fine but the D810 NEF is littered with artefacts. I'm not sure if they're moiré or something else.

Here's the centre circle at 200% - D800E is on the left, D810 on the right. NB. The D800E NEF is lossless compressed because it's only 43.6MB whereas the D810 NEF is uncompressed at 74.2MB. Both have picture control sharpness at +3 in camera and both were shot using manual focus.

[BernardLanguillier]

I agree. The D800E 100 ISO looks better than the D810 100 ISO - it's got smoother detail.

If you blow them both up to 200% and look at the centre circle on the chart, the D800E NEF is fine but the D810 NEF is littered with artefacts. I'm not sure if they're moiré or something else.

The D810 seems much sharper here, doesn't it?

We have typical images with a (weak) AA filter on the left and without one on the right. This is the closest thing to a MFDB I have seen from a DSLR.

Cheers,
Bernard

[Lupin]

duplicated post

[Lupin]

The D810 seems much sharper here, doesn't it?

It does but only at the expense of artefacts. To get rid of them, the D810's in-camera sharpening would probably have to be zero (instead of +3) and that might make it softer than the D800E.

I'd like to test the 800E and 810 myself though - you can never be certain about focussing when looking at someone else's pics.

[Hans Kruse]

I got mine this afternoon. I did some test shots and calibrated it too. A few comparison shots shows to very little difference. It is nice with the EFC, but I don't get the way it is implemented. In live view where this should be the obvious place to use it, I cannot shot continous and I need to set it to MUP and even click the shutter twice for each picture! Compared to the Canon implementation this is really clumsy. The bottom line is that I will live with it. It is nice too that the camera now uses the buffer when shooting in live view!

[Paul2660]

Sadly the dark frame processing is still done without moving it to the buffer so the camera is locked out until the dark frame is done. Canon figured out how to do this years ago.

Paul

[DaveJ]

I believe EFC does work in Live View, not just MU. Pressing the "i" button while in live view brings up the option to activate EFC. It makes sense since the mirror is locked up in LV.
Dave Jolley

[Hans Kruse]

I believe EFC does work in Live View, not just MU. Pressing the "i" button while in live view brings up the option to activate EFC. It makes sense since the mirror is locked up in LV.
Dave Jolley

EFC does work with live view when you also have set the wheel to MUP! Otherwise not.

[fdisilvestro]

Sadly the dark frame processing is still done without moving it to the buffer so the camera is locked out until the dark frame is done. Canon figured out how to do this years ago.

Paul

Dark frame requires another exposure with the same shutter speed as the previous image, I cannot see how you can do this in the buffer since it requires the sensor. After the dark frame is captured, the "processing" is just a simple substraction that does not impact performance.

[BernardLanguillier]

I got mine this afternoon. I did some test shots and calibrated it too. A few comparison shots shows to very little difference. It is nice with the EFC, but I don't get the way it is implemented. In live view where this should be the obvious place to use it, I cannot shot continous and I need to set it to MUP and even click the shutter twice for each picture! Compared to the Canon implementation this is really clumsy.

I wonder how Canon does it. My understanding is that EFC means that the sensor has the ability to stop/start recording information at a certain point of time, right?

This would mean that it must first be told to shut down before being told to start recording. If the order to take the image is implemented on the Canon as both the shut down and shoot orders, doesn't that induce some delay for the start of the actual capture?

But, regardless, the UI chosen by Nikon to switch to this double click capture in live view (need to activate MLU) is somehow debatable, although I don't find to have any impact on shooting experience. It is just something to get used to, isn't it?

Cheers,
Bernard

[Hans Kruse]

I wonder how Canon does it. My understanding is that EFC means that the sensor has the ability to stop/start recording information at a certain point of time, right?

This would mean that it must first be told to shut down before being told to start recording. If the order to take the image is implemented on the Canon as both the shut down and shoot orders, doesn't that induce some delay for the start of the actual capture?

But, regardless, the UI chosen by Nikon to switch to this double click capture in live view (need to activate MLU) is somehow debatable, although I don't find to have any impact on shooting experience. It is just something to get used to, isn't it?

Cheers,
Bernard

I don't believe there is any difference in how Canon and Nikon does it. The EFC works by the sensor doing a reset to start the exposure without the first curtain moving at all. The MUP in live view is not moving the mirror which you can assure yourself about when doing what I suggested some posts above, take off the lens and use a long shutter speed and see what's going on with the shutter. In live view the first click of the shutter does absolutely nothing the second starts the exposure. In fact on the D800(E) what happens in live view with MUP is that the first click on the shutter button closes these first curtain and the second click opens it for exposure. They just kept the same functionality with no reason since the shutter does not move at all with EFC.

On Canon live work works with EFC and you can do continous shooting with only the second curtain moving. That's what I had hoped Nikon would have done with the D810. But not. As I said it's something one can get used to, but ..... it in my view this is a bug in the implementation.