This morning I did the first of 3 different tests I plan on this topic. There have been lots of debates and conversations about color differences between the D8X0 and DF cameras.
I’ll be comparing the D810 and DF specifically. Please note, I have no care on the outcome of this other than finding out for myself what I can from my informal unscientific tests. I have both and plan to keep both cameras.
In my first test (same Ole lake landscape I default to due to its dynamics and detail) the DF metered the same spot 1/10th a second faster than the D810, rather consistently overall with the DF metering for a lighter exposure on a regular basis. [Took 4 Exposures with each camera]
Here are your LR versions, Note, I set the WB and Tint to the same settings for these as there were differences in the RAW import. The same shadow lift and highlight supression was applied to bring these up a bit as they were shot towards preserving the highlights at -.7EV. Both are at ISO 100, F/13. The D810 metered the at 1/50th where as the DF chose 1/40th. The same focus point was used for both and the same lens, Art 35. The D810 set WB at 6100 where as the DF was at 6300 before equalizing them in LR5.
These are OOC with the only settings being changed are those you see in LR above.
Probably more surprising to me after setting the WB and Tint the same in LR the DF had warmer colors. It was no surprise that when I tried to equal them out the D810 leaned towards more saturated colors with a stronger bias to magenta. I have noted this rather consistently since I received my D810 and while the updated RAW conversion for LR seems to have helped, it still seems to lean saturated and magenta biased in my humble opinion.
Feel free to comment yourself on how you see this, that’s why I am making this post.
It would seem though that the histograms don’t display this looking at them below and that the DF has the stronger RED component and even shows more magenta in the lower end of the histogram.
While not displayed here it would seem that raising the WB temperature does not account for the differences. Increasing Vibrancy on the D810 image does seem to move it closer to the DF image.
While the DF image may seem more pleasing to some, I dare do say that on this landscape the D810 does seem to have more accurate color.
If your curious, its obvious that up-scaling the DF image in Perfect Resize doesn’t come close to showing the same detail.. of course if you dont capture it you can’t scale and manufacture what doesn’t exist. You can see that in these crops.
Test two will actually compare a gretagmacbeth color checker card and Test three will compare skin tones in a similar fashion. I realize this test may be flawed, so keep in mind I am conducting it for my own benefit as I believe seeing and testing with ones own eyes and equipment is most beneficial.
The web reduced full pictures are below. No additional processing has been done other than whats been referenced in this post.
These use the new Adobe Profiles and are basically cropped and same sized, no other processing.
The histograms seem much more telling of the differences I have been seeing and the colors are very different with the D810 much more closely matching white. One wonders if that doesn’t influence everything else or not.
Note the difference on Orange and yellow, I can tell you that the DF matches those colors much more accurately looking at the pictures and the color checker. That said the top left blue and gold are better matched on the D810. If you have a gretagmacbeth card you can compare yourself if you have a calibrated monitor.
Finally in TEST 3, (lets concede now that this is a test picture clearly not meant to be commented on artistically;) the skin tones.. they really don’t look too different despite the obvious differences in colors between the cameras on the clothing. Of course the test is limited as all three ladies have similar complexions.
Both pictures were taken in evening light with with Flash, 1/60th F/3.2 Art 35 ISO 100 all Cameras settings as equal as they can be. WB was equalized in LR5 to 6300 and the D810 was downsized for the crops. Yes, the girls are not in the EXACT same pose and position (try keeping 3 young ladies still for 2 minutes:), however, they are in the same spot same light so its much closer a comparison than others I have seen here (minus some I saw of a D800/Df comparison).
A series of three shots were taken to ensure consistency in metering and exposure, there was no difference in the 3 from each camera. They were processed using the new Adobe profiles.
As I said, I think you can see more difference in the clothing colors than the skin. You can see differences in the histograms though, rather interesting.
Shorts are Black, Orange and RED/PINK and shirts are White, Purple & White.
Just in case your wondering, I would choose the DF for portraits as I don’t like analyzing peoples pores and blemishes, not because I think one camera does them better than the other as I would easily grab a D810 to shoot a large group portrait.
This concludes my experiment. My take away was they are both great cameras, no big mystery revealed there.
My DF exposes hotter than my D810 and I am apt to now apply -.3ev in general when I go out or just adjust in PP since I shoot all RAW.
My DF exposes about 200 degrees warmer than my D810. Change in PP if needed.
The D810 colors seem a little more accurate overall, although both cameras could benefit as suggested by custom profiles if color accuracy is a goal or you own both and want to match them for workflow to use both cameras for a job. (I will try a DNG created profile for each and see how close they get). The DF to me still has that magic little something, however, it is getting far less relevant after more time with the D810 and adjusted workflow.
Its worth noting that things can very between cameras and my cameras may behave differently than your cameras and of course this is a very limited set of tests.
Much of the debate and wars people have over these cameras just seems silly in the end, they are both capable and their differences are really very much objective in the end and openly listed in their specs. They are what they are and you either like them, don’t like them or don’t care. So long as you do your homework, try them and understand what your buying they both should serve you well!