Which is better at high ISO: the Canon 5D Mark III or the Nikon D4? Well, the answer to that question has presented itself in a decisive manner: the 5DIII absolutely blows the D4 out of the water at anything above ISO 1600 by at least 1 f-stop. Want the full analysis, go here and read all about it!
So, what does this mean for would-be buyers? Answer: not all that much, yet.
When looking at photographic tests regarding noise, one must remember that the test is being done by looking at the pictures at full resolution, which is something that no one is going to do when it comes to printing pictures on paper or sharing them electronically. In contrast, when printing, most people stop at the 8x11” size, which is the largest most consumer-grade printers go and as big as most people want to pay a pro to print. By shrinking an image that would be about 60 inches wide on a computer screen onto a 8x11 piece of paper, the pixels essentially shrink and, along with them, the noise. Result: a picture that looks noisy when viewed in-full on a computer is silky smooth in print.
The same thin is true when it comes to electronic sharing. Think about it: do you send your friends 12+Mp pictures via email? Doubtful. If you're like most people, you shrink the pictures so that they can be viewed in-full all at once on a computer screen to about 1,000 pixels wide or so. In shrinking the picture, you lose that magnification that shows all of the unsightly noise, meaning that downsampled pictures are always remarkably clean-looking.
Now the caveat: yet.
Two things could make the above information irrelevant: the need to crop/print huge and the next generation of cameras. First thing: cropping. When one crops an image, you are essentially magnifying the portion of it that you are keeping. By zooming in on the view, any faults with the images, like noise, will become more apparent. End result: if you plan on making some heavy crops, expect some noise to become visible due to the magnified view. The same is true of huge printing. If you're a pro planning to print mural-sized images, the noise will show up. If you're an average Joe who will stick with 8x11s, the small size of your picture will erase the noise, promise!
Second consideration: the future. First, there is no doubt that the 5DIII is about a stop better than the D4 when it comes to its image quality. This fact is especially impressive considering that the 5DIII has higher resolution (by 6Mp) than the camera it betters. Clearly, in bettering the 16Mp D4 with the 22Mp 5DIII, the ball is clearly in Canon's court now as Canon has clearly won the high ISO battle for this generation of cameras.
Back in 2007, Nikon leapfrogged long-standing King Canon for the high ISO title. Come 2012, Canon has reasserted its place as the best in the industry in regards to this very important photographic capability. Right now, Canon is better, no doubt, but there is no way to look at the long-term trend until the next generation of pro cameras hit stores, probably in 2015/16. If Nikon fails to close or erase the gap with the D5, then there is considerable reason to possibly consider jumping ship if one is not too heavily invested in the Nikon system, that is unless Canon doesn't offer any real improvement, either.
My advice: for now, stick with what you're shooting as, by what's been outlined above, there's a good chance that the 5DIII's 1-stop advantage is a moot point unless you plan on printing wall-sized pictures or plan to crop a lot.
Still think the D4 sucks? Well, you may want to look at some noise reduction software before dumping your entire Nikon system at a loss and buying into Canon at full-price . After all, as with any product, the winds of fortune change and its simply financial insanity to change systems every time one manufacturer makes a new camera that's a little bit cleaner than the competition. Personally, I'd worry a lot more about the glass I was using than the camera.
For full coverage:Canon announces 5D Mark III
Key specs.
Why did it take Canon so long to listen?
5DIII noiseless through ISO 12,800
Sell your 5DII before no one wants it
5D Mark III vs. Nikon D4 at high ISO
5DIII vs. 5DII, D700, D800, A900
5DIII renders 1Dx irrelevant
Humble requests:
If you found this informative (or at least entertaining), help me pay my bills and check out my Examiner pages for space news, cleveland photography, national photography, and astronomy for more great stuff.
If you think this was cool, why not tell a friend?
For something even better, follow this blog.
Pg 108 28D review
Posts
Umm, not to bust the bubble and point out the obvious - but the images referenced were not shot at the same specifications (e.g. the Nikon 6400 ISO image was F4 and 1/200th while the Canon image at 6400 ISO was F8 and 1/800ish). Beyond the differences in "the controlled" lighting, it is entirely unclear what the processing of the RAW was done in camera and in conversion from RAW to JPEG.
ReplyDelete(Until all this is cleared up, and actually measured by someone like DxO I'll put my money on physics which suggest smaller photo sites will lead to smaller sampling and potentially more noise - this by the way is the reason the D4 doesn't equal the D3s in ISO performance, though it comes within 10%)