Canon EOS 5D Mark II versus Nikon D3

I'm primarily a landscape photographer and prefer the "big vista" type of landscapes that Ansel Adams was famous

for. That being said, I want to move from film to digital. I currently use a Contax RTS-III system to do most of

my work, so that leaves me in the position of not having an existing set of lenses for any current DSLR system

and frees me to choose either Nikon or Canon.

 

There seems to be two primary differences between the D3 and the 5DII that would affect my work: pixel resolution

and build quality. (I don't care about video, so that feature of the Canon is not relevant.)

 

The 5DII is 21MP and the D3 is 12MP. For landscape work where my typical print is either 8"x10" or 16"x20", how

much difference will 9 additional MP of resolution make? If this is a significant factor, then this will probably

sway me towards the 5DII (the EOS 1DS mark III is out of my price range).

 

The D3 is Nikon's top-of-the-line "pro" DSLR. Is its build quality significantly better than an "enthusiast" DSLR

like the 5DII? I'll be carrying whatever I buy around in a rucksack in sometimes dusty conditions--will a pro

model make much of a difference here?

AT 16x20 you'll definitely see the difference. Build quality is not as high in the 5D but I've yet to hear any landscape photgraphers complain about the original 5D in the field, and the II is built better. The 5D also weights less, which can make a difference after a while. Also FWIW, the 5D has finer grain at high ISOs.

Jerry,

 

No question the 5D MkII, the extra 9 mp will be important for fine detail and besides you can get an RTS to EOS adapter for $39 and use all your current lenses, for landscape work the functional limitations shouldn't limit your use.

 

Take care, Scott.

There may be a little bonus here in the Canon EOS line as well. I am not really familiar with the Contax RTS-III, but I presume it uses the Contax/Yashica mount and is manual focus. If so, there are relatively inexpensive adapters that would allow you to use your C/Y lenses on the 5D as MF, stop-down lenses.

 

I bought a Vivitar Series I Qdos lens in a C/Y mount and it works beautifully on my Canon EOS bodies<div></div>

Ditto with the above but if you do decide to go Nikon the D700 would make more sense for your application than the D3.

Sorry, Scott, I somehow missed part two of your answer until after I posted.

Hey no worries JDM,

 

You came up with the picture, as you so often do, the massive telescope/lens is probably my favourite so far :-)

 

Scott......

I don't believe that you'll see any difference from the additional 9 Mpixels if the optical properties of the 5DII are

inferior to the D3.<br>    The 5DII only has 30 percent more resolution than the D3 (5616 x

4256 vs. 3744 x 2832), if you were to compare the print resolution at 20" x 30", the 5DII prints at 187 dpi while the D3

file will print at 142 dpi. If anything, the additional 9 Mpixels can actually hurt the performance of the 5DII and cause

increased noise and lower dynamic range.

Berg Na,

 

That is one of the silliest comments I have seen. The D3 has 12MP times that by 1.75 and you get 21MP of the 5D MkII, ergo 75% more MP, 75% more information must make a difference. The 5D MkII is not an unknown sensor, it is being used and raved about by many landscape photographers in the 1DS MkIII. The detail that the 21MP files give you in landscapes and studios is the very reason it was made. You have obviously never seen or used prints from either camera to make comments like that. Neither the D3 or 5D MkII have any optical properties, that is all down to the lenses. Canon have stated, and the test images seem to support them, that the 5D MkII sensor has better DR and lower noise than the 1DS MkIII.

 

Take care, Scott.

Jerry, I agree with all the posters above, as they make valid points for either system. In particular, the Nikon D700,

since it is substantially lighter than the D3, without sacrificing any image quality (primarily speed, which is moot for

your application), would be a good choice for landscape work, especially with Nikkors such as their awesome 14-

24mm f/2.8, which would yield super-sharp true wide-angle images. Of course, anything said about the supposedly

awesome Canon 5 D MII image quality is now simply speculation, since it's not even available yet (except pre-

production, and I don't put much weight in those evaluations). And IMHO, if you're going to only 16x20 max, I think

that the 12 megapixel sensor of the Nikon D700 is more than adequate, with appropriately good technique. (I've

enlarged *4* megapixel images to 16x20 with very good detail and clarity!) Plus, don't forget ergonomics in your

decision, as they are quite different beasts! Good luck! ~Steve

Scott - Sorry but you're way off: The pixel count is obtained by multiplying the number of rows by the number of columns, so take the square root of your 1.75 factor to get the increase in resolution, or 1.32 (30%). To get 75% better resolution than the D3 you'll a 32.5 megapixel sensor.

Any way you want to put it Berg you can't overlook the fact that if you have 12 million bits of information and I have 21 million bits of information, I have 75% more information than you. But it all boils down to technical semantics, go look at a detailed 20x30inch print from a 1DS MkIII and a D3 and tell me there is no difference.

 

Then if you can honestly tell me I am way off I'll point you to a good optometrist.

 

The rest of my comment I presume you agree with.

 

I am not a pixel peeper or Canon zealot, but the 5D MkII has a tried and tested sensor, it gives superb results, detail from the files is quite simply staggering and as a landscape shooters tool I can't see why you would ever want less detail than you could have. True in 8x10 you won't see the difference. I still use a 4mp camera for some of my work, but I know it's limitations, enlargements of fine detail are best done with the highest number of quality pixels available, if not who would buy a $20,000 and up MF digital back?

 

Take care, Scott.

Steve how can you say "Of course, anything said about the supposedly awesome Canon 5D MII image quality is now simply speculation", the sensor has been available and in use for a good time now, it gives superb results in its current form and I can't believe it will be repackaged in a dumbed down form, that would be commercial suicide for Canon.

 

All reports suggest it is an improvement on the 1DS MkIII, with the extra R&D I can't see why it shouldn't be, the comments you make about the D700 hold up for the 5D MkII as well, it is smaller and lighter than the D3 (and 1DS MkIII) and whilst many images won't show any real difference at 16x20, detailed ones could and the Canon will always give you cropping room, which you will need cos a 16x20 equates to a cropped 16x24 already.

Scott, point taken. You are no doubt much more versed in Canon technology than I. While I do agree that more pixels means more data. (I used to shoot 4x5 and 8x10 sheet film with Schneider and Rodenstock Apo optics, so I'm not unfamiliar with " Image Quality". I think my aim was to suggest that we are talking about an as yet unreleased product, albeit with the sensor being in use already, I think there is more to image quality than the sensor alone...digital converter, processor, lens, and perhaps even the photographer has something to do with it:-). Frankly, I am debating between the Nikon D700 and this Canon 5D Mark II myself, so I am VERY interested in the professional testing comparisons! I've read all the glowing reviews of the Canon 5D's IQ, so I do have high hopes for the Mark II...I just don't know how many pixels you can put into a 35mm frame and still have "Quality"...but I hope for the best! Thanks, Steve

Ah Steve,

 

Now that is a much bigger issue! Obviously there is a good deal more to the end result than the capturing chip, but to be honest I think software programs have reached the level where any sensor/processor combination can be made to imitate any other, including lens colours and even film types, it just takes the interest to search for the results.

 

Lenses are a completely different issue, somebody with your background, I would think, would be far better looking at lens tests for the focal lengths you want to use, another huge advantage for the Canon is its ability to use a large range of non Canon lenses effectively, and certainly primes are the way to go for ultimate quality in landscape and very detailed situations.

 

The pixel count,I suspect, will continue to rise on the FF sensors, diffraction limits have only been reached for the f stops of 11 or so and smaller on full frame. A FF sensor with the pixel density of a 50D equates to 39MP, but it appears there is no practical IQ increase at this density, it is just out resolving the diffraction, this seems to be the factor that is going to limit the 36x24mm sensor, it always was when super fine grained film was used too.

 

Anyway good luck with your decision making too.......

 

Take care, Scott.

Scott - It is not semantics, you simply do not have 75% more resolution. Just compare the print resolution, divide the number of pixel per size by the print linear dimensions, you'll see that the difference is only 30%.<br>

    This common misconception is exactly why manufacturers keep pushing for higher pixel count at the expense of image quality. Cramming more pixels in the same area means a smaller individual pixel size, reducing the dynamic range (less signal capacity), and higher noise due to a lower signal to noise ratio.

I agree with Berg Na on this one. I would take into account the difference between the rows and columns of the sensor,

NOT the total number. So if I want to double the resolution of my 6mp camera, I need to double both axes and get a 24mp

camera and not a 12 mp.

Berg,

 

"Lies, damn lies, and statistics", with respect to Benjamin Disraeli.

 

Believe me I understand, I never said the Canon had 75% more print resolution, I said the Canon produced 75% more information, and it does. You still, obviously, did not go look at a 20x30 inch highly detailed print from a 1DS MkIII, if you had you wouldn't be talking like that. I was a disbeliever for years, I have made 20x30inch prints from a 1D with 4.2MP, I know what extra pixels give you, there is a difference between 12MP large scale detailed prints and 21MP ones.

 

Please go look at some.

 

Take care, Scott.

Scott,

 

If you have two square inches and one of them contains 75% more information which one is going to show more detail?

 

Don't get hung up on the numbers, look at the prints.

 

Take care, Scott.

Scott - I have no doubt that a print from the 21-Mpixel 1DSMkIII will contain more details than another from the 4.2-Mpixel 1D, the former has more than twice the resolution of the latter. However, 30% more resolution will not be so evident. I have no idea what you mean by 75% more "information" if you're not referring to the image resolution.<br>

    It's clear that I won't convince you that the 9 additional Mpixels won't really help, and that's ok. Resolution alone does not make a good photograph.<p>

Best,

Scott Frindel Cole - You are absolutely correct. The following statement is from an<a href="http://luminous-

landscape.com/tutorials/understanding-series/res-demyst.shtml"> article on sensor resolution at the Luminous

Landscape website</a>:<p>

"Images are two-dimensional objects. Therefore, doubling the resolving power necessitates a fourfold increase in

pixel (of equal quality) – not a twofold increase as is commonly believed. All else being equal, a 6 megapixel camera

does not have twice the resolving power of a 3 megapixel camera, but twice that of a 1.5 megapixel camera."

Berg,

 

For goodness sake, stop arguing figures and go and look at some prints!

 

I understand how you get your resolution figure, you don't seem to understand how I get my information figure, but if one pixel gives me one piece of information, two gives me two, it is that simple. I might need four pieces to double my resolution from one piece but two pieces still gives me more detail than the one. 21 million gives me 75% more than 12 million.

 

You don't need to convince me it doesn't make a difference, I have seen the prints and know that it can, I wish you would.

 

Very good point "Resolution alone does not make a good photograph." But superb resolution can make a good picture better. The camera manufacturers can't be blamed for the pictures that are taken with their products, but the best craftsmen make the best masterpieces with tools that do not limit their abilities. 21MP is overkill for 90% of photographers, even most pros, but, the few who can use these numbers, and more, are detailed landscape photographers who make large prints, that would include Jerry and Steve.

 

Take care, Scott.

If you want to have real quality forget the digital, neither Canon or Nikon would do better that film...... if you really want to get the best, move to MF you will see the difference.

 

Aaron

The 5D Mark 2 and D3 are very different cameras. The D3's direct competitor is actually the 1D Mark 3. The D700 is much more of a direct com petitor to the 5D Mark 2, but again they are quite different.

 

The 5D Mark 2 is a high resolution landscape/architecture/portrait camera that's somewhat slower. The D700 has less pixels (larger pixels should allow less high ISO noise, but the 5D Mark 2 seems to be good w/ that also) but features Nikon's top of the line AF module and is capable of 8 fps w/ the battery grip.

Scott,

 

Look at it like this, working on Berg's figures a 20x30inch print from a 5D MkII has 187 dots per linear inch or

187x187 per square inch = 34,969 dots of information, a D3 has 142x142 per square inch = 20,164 dots of

information. 20,164 x1.75 gives a hair over 35,000, the resolution might not have gone up by 75% but the amount

of information contained within the square inch has gone up by 75%.

 

When you get to big prints of fine detailed subjects that 75% more information shows. Fine art printing should

really be done at over 200 dpi, prefferably 250dpi, even the 5D MkII needs to be up scaled a little to do this

but the Nikon needs to be up scaled 100% to make the lowest figure. That means half the information in a

20x30inch fine art print done at the minimum 200dpi from a Nikon file is made up!

 

Hope this clarifies the difference between an arbitrary resolution figure and the more practical amount of

information given figure in high MP files.

 

Take care, Scott.