5dmkIII vs D800 image comparison

[Dylan & Marianne]

I thought this was interesting
Someone's done some comparisons of high iso work between these 2 cameras
Going to do some landscape shooting hopefully this weekend with the mkIII and try to take the same shot with the mkII to see if there's any difference noticeable ( I don't expect much for landscape work!)

http://www.slrlounge.com/on-imaging-...ple-comparison

[LJG]

The Canon/Nikon rivalry has always been there Dylan, just the same as Holden/Ford, but I have never seen it as active, or even as cutting as it is with the release of these 2 models. It seems to be the talking point everywhere with one side claiming victory over the other and vice versa. At the end of the day though we as the consumer are the winners because camera technology is really coming of age now. It makes me wonder what the next few years holds for us.

[Dylan & Marianne]

definitely Lloyd - that's why I thought I 'd do the comparison between mkII and mkIII which I reckon is more meaningful for most since most aren't going to jump camp!

[Bennymiata]

From what I can see at 6400, the Canon wins hands down.
The Canon image is very useable, where the Nikon one is very speckled, and to me, would be unuseable.

Maybe Canon is right.
Bigger pixels = better high ISO quality.

[Darey]

To me these tests would only be meaningful when the same lens is used on both cameras and AF has been micro adjusted.

I'm a firm believer that you only get what you pay for and I don't care which of these two cameras I bought I think the "bang for buck" is incredible for both.

[Dylan & Marianne]

if you ask me, had i been with nikon from the start, I'd have ordered a D800 ages ago! but since I have a long term investment with canon, 5dmkIII aint half bad either

[mongo]

Mongo's view of the results puts Canon out ahead of the Nikon at a number of different ISOs and better overall (and Mongo is a Nikon user). The lenses will play some part in sharpness and contrast but Mongo would not think they can very seriously affect noise.
Must say though that the nikon image was far more magnified and it is pointless sometimes to not use the same focal length of lens or the equivalent resulting focal length of lens on each camera if you are going to do a "side by side" "apples with apples" comparison.

[Tannin]

The incompetent tools have done the comparison using 100% crops! This is idiotic! You can absolutely guarantee an undeserved walkover "win" to the camera with the lower pixel count if you are dumb enough to "test" it that way. Meaningless drivel.

The correct (and indeed only useful) way to compare noise in cameras with identical sensor sizes but different pixel counts is to compare equal portions of the image. You MUST compare like with like, otherwise the "results" are ridiculous.

I can't believe people are still getting stuff as basic as this wrong.

[ricktas]

In the end we are all photographers, and share a love for capturing the world around as withing a fleeting moment of time. Whether you capture that moment with 'the best sensor' or not, is really irrelevant. We use the gear we have, and capture it to the best of our abilities. There is not perfect camera, there will never be the perfect camera (or lens), and whilst the comparisions (good or badly done) are interesting, next year, or the year after a better sensor will come out.

Use what you have, take your photos and enjoy the results. Those who lament the miniscule differences in what a sensor or two can achieve are missing out on the opportunities that awaits them with their present gear. Just cause a new sensor has been released, doesn't automatically make your existing gear bad!

Sometimes we get to absorbed by the technical to remember that the D200 or 40D of years ago, can still take damn good photos.

[jim]

Damned right Rick. Though if I'm going to drop three grand on a camera—and it appears that I am—I'd be quite interested in a well designed comparison study to see how it stacks up against its competitors.

[I @ M]

I can't believe people are still getting stuff as basic as this wrong.

Wow, that is a bit harsh!!!!
That site has done absolutely the right thing with their testing, how can they possibly be wrong when they have a a side by side comparison of the latest and greatest bodies on the net very soon after release --- to start a little controversy and invoke a million more hits to their site each day to keep the advertisers happy.

Tony, you used the word tools to describe them, I really think that is a little to complimentary a term. As you point out the highest res sensor is always going to look worse of in a flawed comparison such as that and if one uses their comparison tool you only have to reverse the scenario and insert a 5D Mk11 and D700 side by side to see similarly skewed results. Those two models have been out for aeons in digital camera terms and both have proven to be excellent bodies for their intended roles just as I am sure the latest models from both companies will perform perfectly for the non pixel peeping measureabators in the future.

Of course their are other points to consider in their "testing" process such as did they shoot raw files and process them identically through the same converter such as ACR?
Somehow I doubt it, the end result images look very much like jpeg shots with default in camera noise reduction and sharpening settings and everyone knows that C and N vary a fair bit with their default profiles.

Tools? Maybe left handed metric shifting spanners.

[arthurking83]

From what I can see at 6400, the Canon wins hands down.
The Canon image is very useable, where the Nikon one is very speckled, and to me, would be unuseable.

Maybe Canon is right.
Bigger pixels = better high ISO quality.

LOL! you got it the wrong way round .. Nikon is right and that bigger pixels = better higher ISO images!
(Nikon have simply hedged their bets now with the D800. D4 for low light, D800 for more detail.)

To me these tests would only be meaningful when the same lens is used on both cameras and AF has been micro adjusted.

...

AF and micro adjustingments will not affect their results.
If Tony(below) is correct and these folks are in fact "tools" and used AF through the viewfinder to acquire the best possible image detail, then the tests can be discounted and deleted from the database without hesitation.
They'd have almost certainly used manual focus .... or alternatively Liveview auto focus where micro adjustingments won't make any difference to AF

.... You MUST compare like with like, otherwise the "results" are ridiculous.

I can't believe people are still getting stuff as basic as this wrong.

This is true for detail in the images where they're trying to show how much detail can be retained from an image, but when it comes to colour accuracy or abnormalities, looking at 100% pixel view is of the utmost importance when analysing the sensors raw performance.
100% pixel view still has it's merits if you know what(or why) you are looking at it.

D800 is certainly bettered by the MkIII, and this will show up even when normalised for a print of equal size, in much better colour reproduction.
But in saying that D800 has nothing to be ashamed of considering the pixel density .... and as Dylan said ... if you were getting into a system(ie. any system) now, you'd be better off going with the Nikon system is money wasn't a major factor.
But as always it goes around in circles .. in a few years time, Canon will find the edge it needs to better Nikon's technology at the time .. just as Canon had the edge over Nikon for many years, many years ago!

The most important conclusion to come out of this is simple .. us consumers win as the manufacturers keep pushing the envelope.
And in he very near future we'll have 100Mp sensors with 20Ev of dynamic range and noiseless ISO500K performance!

[Dylan & Marianne]

wow, what a response to this thread haha - I've just got home from using the 5dmkIII at a wedding without using fill flash and using 3200 iso - will post some later on - the lCD rendered jpegs look great but i suspect the rAWS not as good - will let you know later - I'm not touching the can of worms that is the D800 since i won't be using one ;P

[Tannin]

Not harsh at all, Andrew. Over-kind, if anything.

I have no idea how good or bad the balance of their procedure was, nor do I need to know - as their failure to select a valid fundamental basis of comparison makes all other methodological questions moot. To compare two cameras for noise it is ESSENTIAL to compare equal areas. No ifs, no buts, no maybes. This is the essential basic equality without which comparisons are meaningless.

[TEITZY]

Yes in practical terms, if you are printing these images at the same size then the D800 noise will look better than it does at the pixel level. Also these are jpegs so likely some in-camera NR happening which can throw things out. The interesting thing to me is that the exif says the D800 is 2/3 stop faster than the 5D3 (1/1250 vs 1/800). If this holds true across the comparative Nikon & Canon lenses this is significant for low light handheld shooting.

Dylan I would also be interested in a comparison of the 5D2 & 5D3 using the same the lens with identical settings to see if the exposure is the same for each camera.

Cheers
Leigh

[Allann]

I don't mean to be rude, and this is not an attack against you Tony, but I appreciate the effort put into these, and not being a purest can appreciate them for what they are, an indicator. How about showing us some of your photos again instead! They were awesome and some of the best I had seen and kept me striving for the same in my work.