1,086 of 1,202 people found the following review helpful
Unbiased Review of 5DmkIII vs D800,
This review is from: Nikon D800 36.3 MP CMOS FX-Format Digital SLR Camera (Body Only) (OLD MODEL) (Electronics)
I'm a Sony shooter with only a few lenses. I use to shoot Canon during the film days. I tested both the Canon 5D mkIII and the Nikon D800 and here are my results. I tested a 5DmkIII with a 24-70 f2.8 lens and a Nikon D800 (not D800e) with a 24-70 f2.8 lens.
About the same, except D800 has a lot more detail to work with. In Lightroom, I can save a higher percentage of ISO 6400 shots because the D800 has more detail. Canon seems cleaner initially in Lightroom but when the picture fits into a 24" 1920x1080 monitor or a 64" Samsung plasma TV, the Nikon looks a tad cleaner, noise less noticeable. I think the Canon looks cleaner in Lightroom because it is just a smaller picture. But displaying ISO 6400 shots on a monitor or TV, Nikon looks nicer in general. Both cameras at ISO 12,800 look awful and not recoverable in Lightroom. It might look acceptable as a really small pic but why the heck would you buy an expensive camera to display crappy looking pictures?
Frames per second:
Easy winner. Canon can shoot 6 FPS, Nikon 4 FPS. However, in practice I think 4 FPS is pretty good. None of these cameras are really Olympic style sports cameras.
Easy winner. Nikon's RAW files are more detailed, almost 3D like. I can't really explain it other than the pictures look more real. I can crop a photo to 1/3 it's size (12 megapixels) and it still looks stunning. I wonder how much better the D800e version is. I'll have to wait until my friend receives his to find out.
Easy winner. Out of the camera, the Canon JPEGs are phenomenal. The processing done is quite remarkable.
Easy winner. If you shoot RAW, Nikon is it. Also there is an issue with the Canon with the color red. I think the color is overblown at times because all the details are loss and not recoverable in Lightroom. Not always but it has happened at least twice. The same photo on the Nikon kept all the details.
About the same. Canon and Nikon have awesomely quick autofocus and I couldn't determine a difference. The only caveat is that Nikon focuses better in lowlight (without the autofocus assist lamp) and also the Nikon focuses when there is almost no light (with the autofocus assist lamp). Why the heck doesn't Canon include an autofocus assist lamp is beyond me. Also, Nikon's face detection is extremely useful because it focuses right on the eyeballs.
Easy winner. Canon doesn't have built in flash. Nikon flash worked surprisingly well.
Easy winner. Nikon knocked it out of the park. I got a lot less blown highlights with shots with white clothing and more realistic blue skys. Also, there is a lot more headroom on both the highlights and shadows on the Nikon when editing in Lightroom. How did Nikon have better shadows and highlights! They have to share some of that technology!
Suprisingly about the same. I would have thought that Canon's lead in video would maintain. Surprisingly, the Nikon's video was just as awesome as the Canon. Nikon's video has a bit more detail and is definitely a little sharper than the Canon. I didn't test Nikon's uncompressed HDMI out, although it seems to be a useful feature (this is like RAW HDMI output for video). Canon should adopt uncompressed HDMI out also.
Easy winner. Why anyone (who doesn't already have Nikon or Canon lenses) would buy this Canon for $500 more than the Nikon would need to think twice. I can see why the Nikon is selling so much better than the Canon, at least on Amazon.
Both about the same weight. Both feel nice in the hand. Canon possibly slightly more comfortable if you have bigger hands. Nikon maybe more comfortable with smaller hands. Both are fine though.
Nikon wins with USB 3.0. Skipping the card reader altogether by just plugging in the camera to the computer is convenient. Also the transfer speed is much much faster than Canon's older USB 2.0. This saves a lot of time.
Canon has slightly better screen in direct sunlight (LCD facing up towards the sun). Nikon is better when the LCD is not directly facing the sun. Nikon's screen is crisper and more 3D like.
My last day of shooting was in a light mist/drizzle. I was shooting both cameras again for about 20 minutes when the Canon 5D mkIII developed some fogging inside the viewfinder screen. I could not wipe it away as it seemed to be inside the camera. I could no longer take pictures normally without live view. Nikon didn't have this problem and I continued to shoot the rest of the day with the Nikon in the same wet conditions without issue. I had planned to shoot at least 2 weeks with both cameras so this was definitely a bummer.
Both Canon and Nikon's 24-70 2.8 lenses are great. I would say the Canon 24-70 2.8 is just a tad faster on focus. Nikon is slightly sharper in the corners. Both Nikon and Canon seem to have a very comparable lens assortment (although my wallet won't be happy buying so many new lenses!)
Well, after using both cameras for about a week, I kept the Nikon D800 and returned my (possibly water damaged) Canon 5D mkIII. Both are phenomenal cameras but D800 has definitely outclassed the 5D mkIII in this round. Maybe Canon will come back strong with its next version. It definitely has some catching up to do.
Thanks for reading my review and I'll also post this on the Nikon D800 and Canon 5D mkIII review page. Best of luck to all you photogs and enjoy these phenomenal cameras!
Sort: Oldest first | Newest first
Showing 31-40 of 82 posts in this discussion
In reply to an earlier post on Apr 16, 2012, 6:58:43 PM PDT
Last edited by the author on Apr 16, 2012, 7:02:34 PM PDT
In reply to an earlier post on Apr 16, 2012, 7:19:23 PM PDT
Last edited by the author on Apr 16, 2012, 7:43:26 PM PDT
We're generalizing? I thought it was "discussing precision topics, then comprehend precisely". OK then I'll disregard the fact that at the distance your audience will be viewing Cindy C on the face of Hoover Dam, 6 mp will probably supply all the detail they can make out. Oh well no matter...
Yeah man I'm cool with that. What you said. I can't blame anyone for wanting the very best there is to offer. But some may want to take a look at my comparisons after both up-sampling and down-sampling identical images made from the D700, D800, and 5D Mark II before fully buying into the mega-pixel myth. These are 100% crops, which means the entire 36 MP image viewed at this magnification would be 6 x 4 feet. But feel free to view away at 1600% ;)
First the "blind test", Canon 5D Mark III up-converted to 36MP vs. D800 native. Can you tell which is which? Yes, there is a slight difference, but more importantly, is the difference significant at 6 x 4 foot?
Now, the full set comparing D700 and D800 at both 12 and 36 MP, and D800 vs 5D Mark III both at 22 and 36 MP. Be sure to select "view all sizes" and "original size".
RAW images courtesy of www.imaging-resource.com
In reply to an earlier post on Apr 16, 2012, 8:48:24 PM PDT
Last edited by the author on Apr 16, 2012, 8:49:10 PM PDT
In reply to an earlier post on Apr 16, 2012, 10:10:26 PM PDT
Last edited by the author on Apr 16, 2012, 10:22:53 PM PDT
"How do you up convert an original image of 21 megapixels to 36 megapixels?"
You don't know how to resize an image? Enlarge an image? It's very easy and is a basic function to anyone familiar with Photoshop or any other image editing program.
The simplest way is to use Photoshop's "Image Size" command with bicubic sampling. Simply enter the dimensions and Photoshop uses its bicubic algorithm to enlarge the image, re-sampling to average out pixels. However there are fractal-based programs which do a slightly better job at enlarging. Re-sampling the smaller image up to the size of the larger one allows you to compare them at the same megapixel size so you can actually see if there is more detail in one or the other. Larger mega-pixel images don't necessarily contain more detail information, as you can see in the side-by-side comparisons.
"You don't even have 2 separate images there. I checked your image online and there's only one image. What are you doing?"
There are two image crops, you will see one on the right, and one on the left. One is a 100% magnification crop from the 36 MP D800 RAW image and one is a crop from the 5D Mark III RAW image that has been enlarged to the same pixel dimensions of the D800: 7360 x 4912 pixels. The two cropped images are just a small section from the entire images which would be 6 feet wide at this magnification. They were placed side-by-side and the two layers merged and saved as a highest quality JPEG for the purpose of posting on Flickr. I'm not sure why that's hard to understand but those are the specifics.
"why do you take two words out of my sentence, (generalize and precision), and compare them instead of comparing my sentences."
OK your whole sentences word-for word were as follows. Your very first sentence in addressing me read: "Ya know, the problem with people today, or maybe ever, is that when they are discussing precision topics, then comprehend precisely."
So, when I responded next, I used more precise language... and then your very next post began: "Awright, awright! Let's get on track. I know you like to cite specifics but we're generalizing here."
Anyway, in an effort to "get back on track of the original intent to evaluate two $3200 cameras", I downloaded RAW images taken under very controlled studio conditions by the folks at image-resource.com, processed the RAW images using identical settings in ACR, resized them to the same dimensions, grabbed some 100% crops from the same area of the photo, laid them side-by-side and posted a set on Flickr with all the results. I would be interested to see if anyone can tell the unlabeled images apart; there are more labeled images in the Flickr set at the second link I posted.
In one test I re-sized the 36 MP D800 image DOWN to match the resolution of 1) the 12mp D700, and 2) the 22mp 5D Mark III.
The second set of tests was more challenging to the lower MP cameras as I enlarged both the D700 and the 5D Mark III UP to the size of the D800 images, 7360 x 4912.
The comparisons are all in the Flickr set and except for the "blind test" are all labeled.
I hope this clears up any confusion.
In reply to an earlier post on Apr 16, 2012, 10:51:26 PM PDT
Deepak Chourasia says:
You enlarged and sharpened.. Do the same thing with d800 photo and then compare... If it is about matching the size then at 80mp enlarged file d800 will show much better details and sharpness.
In reply to an earlier post on Apr 17, 2012, 3:21:53 AM PDT
Last edited by the author on Apr 17, 2012, 3:53:21 AM PDT
No Deepak I only enlarged with Photozoom, a high-quality enlarging program, and I did not sharpen before or after enlarging... But here is what you requested, both D800 and 5D Mark III enlarged to 80mp. In this case the D800 is enlarged about 2.2x and the 5D Mark III is enlarged about 2.8x for them to both reach the same pixel dimensions of 16356 x 10904.
This time I just used simple enlargement using bicubic sampling, no sharpening on either image before or after resizing. As you can see there are some areas where the D800 has slightly finer detail but even at this very high level of magnification, certainly not the "much better details and sharpness" that you were expecting. And there are some areas, particularly in the cloth sections, where the 5D Mark III is actually revealing more of the pattern details.
At this level of magnification, the full images would be 13 feet across! At a more "normal" poster size of 36" across or less, any of the slight differences between the 22mp and 36mp images would be undetectable. Some modest sharpening of the 5D Mark III image to bring out a little more detail would make these crops even more alike, but as I said I didn't do any sharpening for this test, only some minor color balancing to match the colors.
In reply to an earlier post on Apr 18, 2012, 9:46:38 PM PDT
Last edited by the author on Apr 18, 2012, 9:57:47 PM PDT
In reply to an earlier post on Apr 19, 2012, 2:19:18 AM PDT
Thank you for your comments. Comparing images at 100% is fine if both images are the same number of megapixels, however a 100% view of a 36mp image is zoomed in to a much smaller area than a 100% view of a 12mp image. Upsampling the 12mp image to 36mp using Photoshop or a good fractal enlarging program puts both images on an even playing field for zooming in.
This is the same technique that Ken Rockwell used when he did his comparison of these cameras. To quote from his review:
"The D800 images are shown at 100%, while the other cameras were upsampled to the same size in Photoshop CS5 (bicubic) to ensure a fair comparison."
The only problem I have with his comparison is that he used JPEGs right out of the camera, while I used RAW images that were processed in ACR to optimum image quality. In my test the 5D Mark III fared much better, while the D800 still holds a very slight edge in resolution (that can only be seen if the image is produced 6 feet wide!!)
The 5D Mark III and the D800 are both marvelous cameras. I didn't intend any criticisms of you through my remarks. I think you should by all means save up and get the D800. Go for it -- and happy shooting!!
In reply to an earlier post on Apr 21, 2012, 8:56:38 PM PDT
Last edited by the author on Apr 21, 2012, 8:59:41 PM PDT
In reply to an earlier post on Apr 21, 2012, 10:33:06 PM PDT
Last edited by the author on Apr 21, 2012, 11:32:06 PM PDT
Yes I'm familiar with this guy's videos. And he confirms exactly what I was saying:
"In theory you should get more noise with more megapixels. Yes, when you look at them in the 100% crops you will notice that the D800 is a little bit more rough compared to the D4. But when you look at the 100% crops of the D800, you're looking in a lot closer than the D4 image. Look at the images side by side at equivalent magnifications so that both of your subjects are at the same size, and the difference is negligible."
What my testing did is exactly that: enable the viewer to look at the images side by side at equivalent magnifications so the subject is the same size. And when I do that I don't see a real significant difference in either noise OR detail resolution between the 5D Mark III and the D800, certainly nothing to support the idea that "pixel count is the most important factor in photo quality", no matter which method you use to compare the images. Looking at all 3 of these cameras, the 5DM3, the D800, and the D4, the 5DM3 turns out to strike a very nice balance between resolution, speed, and price.
Either these cameras are capable of creating great photos; unfortunately it's somewhat academic at this point since since there are none available to actually purchase anywhere. In a few months when the D800 is fully available you will probably see some Canon shooters selling off all their equipment and jumping ship to Nikon. I'm not sure if that's such wise move in the long haul since these companies seem to leapfrog every few years.
I do know that for me the significant advancement in dynamic range in the D800 is a much more compelling reason to consider switching than the difference in megapixels, unless your application is to print mural size prints ten feet wide.