D7100 vs D7200 at ISO 6400

[Barry Clemmons Photography]

<p>I ran a test tonight using the D7100 and D7200, both at ISO 6400 to see if there is an improvement with the new camera. Both shots were with the same lens (24-70mm) and settings made on the same tripod. Both shots are jpegs straight out of the camera and cropped at 100%. To me there is a noticeable difference, but others may not see what I am seeing with the full size files. The shot below is from the D7100.</p><div></div>

[Barry Clemmons Photography]

<p>Here is the same shot using the D7200.</p><div></div>

[User_6502147]

<p>Would like to hear your analysis. Yes, the D7200 has less noise, but there seem to be more contrast in the frame, as well. Hmm, considering 7000/7100 and 7200, what is the actual DR jump...it's hard to tell. Anyway, it looks really good IMO for ISO6400, especially at 100% crop.</p>

<p>Les</p>

[pete_s.]

<p>The D7200 looks like a big improvement. Examining it closer it doesn't look like it's only better software noise reduction as you have more detail as well in areas like the dog.</p>

<p>On thing to watch out for though when making comparisons is to get the focus absolutely spot on the same between the comparison shots.</p>

[Barry Clemmons Photography]

<p>I used the same focus point with both cameras to acquire focus. <br>

By the way, there is one difference between the two bodies if you use back body focus like I do. I assigned the AE-L/AF-L button to focus when pressed, but a half press with the shutter release also focused the camera. I had to read the manual (drat!) to discover a new menu item (a4) that disables the shutter for AF activation.</p>

[ShunCheung]

<p>Interesting result. Thanks Barry.</p>

[sjmurray]

<p>I'd like to see a comparison when shot in raw and converted to jpg using identical parameters.</p>

[mag_miksch]

<blockquote>

<p><em>I'd like to see a comparison when shot in raw and converted to jpg using identical parameters.</em></p>

</blockquote>

<p><em><br /></em>Same here, OOC jpegs are the result of setting camera software. <br>

<em> </em></p>

[richard_driscoll]

<p>+1<br>

The 7200 will probably have some local contrast enhancement (they call it clarity) set by default and it is not an option on the 7100. I think I can see the effect of that, so pretty meaningless without the same jpeg processing.<br>

I'm amazed by the effect of modern jpeg processing on the raw images from my old D40. With capture NX-D all the options are available except active D lighting, which I think may be because of royalties.</p>

 

[Peter_in_PA]

<p>I wonder if it would make any difference when the image is finally used (whether printed or on-screen)...</p>

[ShunCheung]

<p>According to the one A/B comparison shown here, the D7200 image is clearly cleaner, and that will show up in the final image used unless all you care about is a tiny image shown on the web or e-mail, such as forums like this one, FaceBook, etc. A tiny JPEG can hide a lot of issues.</p>

<p>While that is encouraging, I wouldn't draw any conclusions based on merely one image sample. We need to compare the D7100 and D7200 under various lighting and ISO settings. In particular, we need RAW converters for the D7200 from various sources.</p>

[Peter_in_PA]

<p>that is a 100% crop, though, Shun.</p>

<p>I bet if it was printed at 8 x 10 or smaller, you wouldn't see a difference, even right from the jpeg.</p>

<p>And if both were shot raw and some judicial post processing was done, they'd be more or less identical I suspect. Your last point is spot on.</p>

<p>Years ago, when I upgraded from a D50 to a D90, I did a rigorous comparison at high ISO.</p>

<p>100% crops were obvious, I got 2 and a half stops "less noise" when you got to high ISO.</p>

<p>When I did test prints "normally", the difference was way way smaller. I am confident that would be the case here.</p>

<p>We measurebate this stuff so much on forums like this (and I think you should measure. When I get a new camera/lens, the first thing I do is rigorously test it and compare to whatever I can), but often we are shooting on the fly, hand-held, or making "smaller" prints, and things like high resolution and sometimes even high ISO (some of my favorite images are too noisy for some of y'all) disappear in the face of a great image.</p>

[ShunCheung]

<p>Peter, the average Jane or Joe off the street might not see much of a difference, but I think I'll see a difference in an 8x10 print. That is the "my customers cannot see any difference" routine that is frequently used on this forum; it merely highlights the fact that those customers are not critical photo editors.</p>

<p>The two samples Barry posted are showing a lot more difference than what I had expected from a D7100 to D7200 upgrade, which I thought would be minor. But again, we need a lot more test samples before we can draw any conclusion. There is always a chance that Barry might have messed up the D7100 sample, for example. :-)</p>

[Andrew Garrard]

<p>I'd expect JPEG to show a difference - Nikon have been tuning their engine, as seen in the D4/D4s and D800/D810 generations, which have very similar sensor behaviour but a bigger jump in JPEG. (You could argue the same for the 5D2 vs 5D3 in Canon land.) Of course, if using raw and running DxO's PRIME over it afterwards is an option, you can recover a lot. There's something to be said for JPEGs if you're trying to use a D7200 as a fast camera, though, so I'm vaguely interested to see it - for the D810, I really don't care what the JPEGs look like (although the on-LCD previews were alarmingly good at ISO 12800).</p>

[ShunCheung]

<p>There is a risk for mixing up all these different situations. The D4S has an improved sensor from the D4. While I haven't used the D4, I have tested the Df which shares the same sensor. The D4S is definitely an improvement from that.</p>

<p>However, I find the high-ISO performance between the D800 and D810 very similar. There is hardly any improvement on the D810 @ ISO 6400.</p>

[Barry Clemmons Photography]

<p>I also have the raw images from the above two shots and will do my normal processing on them later today. As info, I also made two shots at ISO 12800 for comparison while I had everything set up. Like Shun I was surprised by the difference in the two cameras at 6400. I feel certain it is due to tuning by Nikon on the D7200 and makes one wonder if the same could be done for the D7100 through a firmware upgrade.</p>

[elliot1]

<p>JPG quality seems to have quite a bit of improvement from generation to generation. But RAW files tell the real story. Remember just a short while ago with the fuss about the D750 Vs the D610/D600. In the end, the JPG images showed marked improvement but the RAW files where basically the same. I suspect the same will hold true between the D7200 and D7100.</p>

[anthony_decarlo1]

<p>First, thank you for taking time out to post, it's always good to a get first hand opinion and hard information from a user with photos to compare. Although to me the dark watch band is very telling, I tend to agree with another post or two, we can't tell to what extent this is due to updated in-camera processing. Sorry to be asking but if you could provide a crop of the raw images <em>without any processing</em> except the cropping, that would be great. If you would like the crops copied and returned after running DxO prime, I can run that for you.<br>

Thanks in advance, Tony</p>

[Barry Clemmons Photography]

<p>For comparison, here is the D7100 shot at ISO 12800.</p><div></div>

[Barry Clemmons Photography]

<p>And here is the D7200 at 12800.</p><div></div>

[Barry Clemmons Photography]

<p>Tony, are you saying you want the photos cropped and saved as raw files for posting?</p>

[luke_kaven]

<p>A number of us were watching to see if Nikon would insist that Toshiba up the performance of their sensors, which had difficulties with pattern noise in the D7100 model. I'll mention again Bill Claff's recent tests using stacked dark frames, which do show a significant improvement in the D7200, enough to explain the observed differences in the results posted here. I really do not think that this is due to improved DSP or firmware, which would be a very poor way of solving a problem that should be solved in hardware.</p>

<p>http://www.dpreview.com/forums/post/55513400</p>

[Andrew Garrard]

<p>As Shun pointed out, sorry if I was confusing in the D4/D4s case. The D4s tests better at high ISO (and the D4 slightly better at low ISO) - but it's quite a small improvement. The JPEG improvement is greater. Likewise, the 5D3 <i>is</i> better in low light than the 5D2 - but not by as much as the JPEGs make it appear. Apologies if I spread misinformation. But cool to know about the D7200 improvements, especially the pattern noise thing.</p>

[Barry Clemmons Photography]

<p>Shun, please remove my last two posts since I just noticed that they are not at 100% crop. I was using NX-D and am not familiar with it yet, but had to use it since it is the only NEF converter available at present. I will re-do them and post again. Sorry about that.</p>

[Barry Clemmons Photography]

<p>Okay, I'm not a huge fan of NX-D but it is all I had to work with on the raw files. Here is the D7100 at ISO 6400 raw file converted to jpeg with no processing other than sizing for the website.</p><div></div>