fredmiranda.com Login
This topic is in the 2014 archive. You can read it and like its posts here. It is read only.

Archive 2014Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)

  
 
isaacimage
Offline
Details
p.2 #1Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
mmurph wrote:
This is only tangentially related to the test above, but it is interesting that the 1DX did hit the focus.

One thing that I have never gotten a clear answer on, is the accuracy of the auto focus on various Nikon systems.

We know from Roger Cicala's tests that the Canon 1DX and 5DIII have more precise auto focus than most other cameras. Most Phase Detection systems have some error:

http://www.lensrentals.com/blog/2012/08/autofocus-reality-part-3b-canon-cameras

http://www.lensrentals.com/blog/2012/09/autofocus-reality-part-4-nikon-full-frame

As recently as February, Roger has implied that those are still the only 2 cameras with more precise Phase Detection. I am surprised that we haven't heard more about this:

http://www.lensrentals.com/blog/2014/02/why-you-cant-optically-test-your-lens-with-autofocus

I know that I
...Show more →

Thank you mmurph will definitely take a look on it.
Looks like great articles.
Especially if it's from Roger
Aug 12, 2014 at 09:29 PM
form
Offline
Details
p.2 #2Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
Was the lens calibrated on the D810? Was the focus point on the right spot?

It's really hard to infer specific image quality differences or AF accuracy from this "test."
Aug 12, 2014 at 10:01 PM
Pixel Perfect
Offline
Details
p.2 #3Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
bernardl wrote:
I see sharpening halos in the 1Dx file, none on the D810.

Besides, I don't think you focused the D810 as well as it could have been.

But still, very nice file from the 1Dx.

Cheers,
Bernard


I see the 1DX as focused on the eye and the D810 has focused on the eyebrow. Always use LV to focus manually when doing this kind of test IMO. I'd like to see also a comparison of 1DX + 85 + 1.4x to D810 + 85 and then look at 100% crops to see how close detail is when features will now be same size.
Aug 12, 2014 at 11:56 PM
mmurph
Offline
Details
p.2 #4Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
isaacimage wrote:
Especially if it's from Roger


There are 2 additional, preliminary articles in that series that you can go back and read. Roger does a great job statistically demonstrating the variance.

It would be great if he went one step further, and created a measure of the error in Phase Detection auto focus. Maybe as standard deviations from perfect focus as a ratio of the lenses depth of field wide open?

With a few standard lenses at standard distances. Or even just the best 50 mm 1.4.

Otherwise we are left without a meaningful way to describe how accurate the auto focus system is from camera to camera. Other than knowing that the 1DX and 5D3 are excellent, and are the best out there.

(He did some earlier tests with Canon bodied that should a sample of those variances and where they fell, which is helpful.)

I'll try to think up a cool name for this metric. The "Roger Cicala Phase Detection Autofocus Error Standard Deviation as a Percentage of Depth of Field at Widest Aperture at 50x focal length"

The "RCPDAEPDFWA50FL" index?? Not quite? How about the "Slop Index" ?

Cheers!
Michael
Aug 13, 2014 at 12:04 AM
dhphoto
Offline
Details
p.2 #5Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
This test has very little point IMHO

However what I think it does show is that file size isn't everything and the actual differences in the real world between any of these cameras is very small (small enough to nitpick tiny details), they are all very very good.
Aug 13, 2014 at 01:06 AM
justruss
Offline
Details
p.2 #6Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
This test does show us that: when you focus on the eyeball, with a lens stopped down a bit, you get sharper eyeballs compared to a different lens wide open that is focused on the eyebrow on a different camera.

I think part of the pushback to this "test" is that the title of the thread is "Canon 1dx vs. Nikon d810"-- strongly implying that you're testing the two cameras. What this test is really showing is two different lenses shot under different aperture regimes (wide open vs. stopped down) and potentially focused on different points (we don't know if the lens AF blew focus, the OP MF'd on the wrong thing, etc).

As others have mentioned, you probably want to do this test from a tripod, with a shared lens (say Sigma 85 1.4), at identical apertures (if you want to test camera sensor, stop it down a bit, even to f/2 or f/2.8), and controlled or semi-controlled light, using live-view MF with a best-of-X approach (x=3 is probably fine).

I would say that there's just so much going on in the test posted here that we can't really draw conclusions about anything individually, like camera, lens, light, user error, etc. It might show us something about skin tones, but even that is a bit overdetermined by a number of factors.
Aug 13, 2014 at 01:10 AM
Gust
Offline
Details
p.2 #7Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
Want to see a landscape or picture with much more details please.
close detail on a smal tele doesn't matter that much, but what happens in the same test with details on distance.

thnx
Aug 13, 2014 at 04:02 AM
BenV
Offline
Details
p.2 #8Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
This test is great. It shows that two cameras can both take a photo. Other than that this was pointless. Your next test should be a cell phone photo of a chair vs. a 3rd graders drawing of that chair.
Aug 13, 2014 at 07:43 AM
brian_sp
Offline
[X]
Details
p.2 #9Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
BenV wrote:
This test is great. It shows that two cameras can both take a photo. Other than that this was pointless. Your next test should be a cell phone photo of a chair vs. a 3rd graders drawing of that chair.




one of the dumbest tests i have yet to see

Aug 13, 2014 at 06:31 PM
EB-1
Offline
Details
p.2 #10Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
OMG, what is the point of this? Are we comparing oranges to pomegranates now?

EBH
Aug 13, 2014 at 06:36 PM
jcolwell
Offline
Details
p.2 #11Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
This thread is loaded, but I'm not.
Aug 13, 2014 at 06:50 PM
Thorsten
Offline
Details
p.2 #12Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
isaacimage wrote:
Hello dear friends
I'm shooting people mostly and was very interested in this comparison,
So for those who is also interested here we go


Thank you for sharing this! And I would not worry about the negative comments. It looks some folks will require a thesis project to be done before any comparison can be posted, and even then would probably find issues. I do those quick and dirty comparisons like yours all the time to get an initial feeling how certain combos work.

Aug 13, 2014 at 06:55 PM
johnctharp
Offline
Details
p.2 #13Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
Thorsten wrote:
Thank you for sharing this! And I would not worry about the negative comments. It looks some folks will require a thesis project to be done before any comparison can be posted, and even then would probably find issues. I do those quick and dirty comparisons like yours all the time to get an initial feeling how certain combos work.



Sure- but if you're going to compare something, minimize the variables.
Aug 13, 2014 at 07:13 PM
dehowie
Offline
Details
p.2 #14Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
Mmm i forgot we shoot portrait work from tripods using Sigma lenses.
Better leave the 85L and G in the cabinet.
Wtf would you test two cameras with a third party lens and all of the AF issues that can cause?
You test with the brand lens they will use in the real world.
This allows both systems the best chance to show what they can do in the "real" world.




Aug 13, 2014 at 09:13 PM
BenV
Offline
Details
p.2 #15Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)


dehowie wrote:
Mmm i forgot we shoot portrait work from tripods using Sigma lenses.
Better leave the 85L and G in the cabinet.
Wtf would you test two cameras with a third party lens and all of the AF issues that can cause?
You test with the brand lens they will use in the real world.
This allows both systems the best chance to show what they can do in the "real" world.




Real world testing is much more opinionated vs. hard facts.
Aug 13, 2014 at 09:24 PM
EB-1
Offline
Details
p.2 #16Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
There are a lot of different real worlds. It's just an odd comparison. The 1D-X is much more like the D4/D4s than anything.

EBH
Aug 13, 2014 at 09:33 PM
billsnature
Offline
Details
p.2 #17Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
I am not surprised by this result. Maybe that is because I own a D800E and a 1Dx.

While I agree the 1Dx is more directly comparable to the D4(s), there is something about the color, contrast and "micro-contrast" of the 1Dx that make it look much sharper and finer resolving than you would expect based on the small MP step up from the D4, or the major "deficiency" compared to the D800E/D810.

I really do like the look of 1Dx files converted from raw with minimal PP. But then I don't shoot photos of people on purpose, and when they end up in photo, I clone them out.

Bill
Aug 13, 2014 at 10:19 PM
hijazist
Offline
Details
p.2 #18Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
dehowie wrote:
Mmm i forgot we shoot portrait work from tripods using Sigma lenses.
Better leave the 85L and G in the cabinet.
Wtf would you test two cameras with a third party lens and all of the AF issues that can cause?
You test with the brand lens they will use in the real world.
This allows both systems the best chance to show what they can do in the "real" world.




Mmmmm... ok, so instead you choose a $500 consumer lens on a 36mp camera and use it wide open and compare it against a $2000 lens stopped down from 1.2 to 1.8, is that your real world? Biased much?

OP, since you have both, why not stop both down to 2.5 (forget 5.6 like I asked before). Even better why not shoot both wide open?

Is asking to have at least some objectivity means we're negative?
Aug 13, 2014 at 10:31 PM
skibum5
Offline
Details
p.2 #19Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
something looks wrong, the D810 shot looks light something went wrong when you took it, it's all blurry (when some D810 shots I see are kinda too sharp what with the AA filter lacking and all)
Aug 13, 2014 at 10:46 PM
justruss
Offline
Details
p.2 #20Canon 1dx vs Nikon d810 for portraits in non scientific environment ;)
dehowie wrote:
Mmm i forgot we shoot portrait work from tripods using Sigma lenses.
Better leave the 85L and G in the cabinet.
Wtf would you test two cameras with a third party lens and all of the AF issues that can cause?
You test with the brand lens they will use in the real world.
This allows both systems the best chance to show what they can do in the "real" world.




The problem is this wasn't called a "test of two portrait setups I happen to have and use when I shot them together once" but a test of the "1Dx vs. D810."

Do you see the issue?

What we're saying is that all the variables involved make it very difficult to draw any conclusions about the cameras, specifically, from the test conducted. In fact, it's difficult to draw conclusions about the hardware so much, since there is enough variability in shooting (read: potential human error/misfocus/vibrations/shaky hands) to even draw serious comparisons between the hardware setups themselves.

There are, of course, multiple ways to test/review camera hardware. We have the near totally sterile, objective measurement approaches (though people will still argue about those), often, in my opinion, exemplified by Roger Cicala, because he even looks at variation among the same lens to give a better distribution of possible outcomes. And we have the stuff on the other extreme: the "I don't shoot test charts and I don't shoot comparable images... I just go out and shoot and say what I see in the using and results."

Both approaches are valid, particularly if we take them together to draw a more complete picture, and particularly when we have multiple outlets and individuals doing similar tests.

The test in this thread is a little too much of both approaches in one, so that nobody is going to be satisfied in either direction. It's not controlled enough for an objective test, and it doesn't cover enough "real world" out in the world shooting examples, or discourse on these aspects of shooting, to give us much of an idea of the OP's subjective/reported-use findings.

At the same time: I find some of the responses in this thread fairly rude. Piling on the OP, and posting things like "Stupidest. Test. Ever." doesn't help this community foster a friendly environment (for new or veteran members), and help others interested in conducting tests get an idea of how to conduct such a test in a more useful manner. Hell, it might prevent people from conducting tests at all-- if mistakes are only to be responded to with mirthful mockery instead of constructive criticism or discussion.

Aug 14, 2014 at 01:18 AM
FM ForumsCanon ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister