« JPEGophobia | Main | Random Excellence: David Dewhurst »

Wednesday, 14 March 2012

Comments

The D800 should "blow away" the 5DMKIII but using their comparator, it doesn't. In fact the MKIII ISO 100 still life side-by-side with the D800 looks significantly better to me.

"The D800 should "blow away" the 5DMKIII but using their comparator, it doesn't."

Not where I'm looking. The fabric samples in the iso 100 D800 image look like fabric and the 5DMKIII does not. As a matter of fact the iso 1600 D800 fabric detail holds up way better than the iso 100 5DMKIII.

I'm a little dubious about the sharpening and compression of the jepgs though.

Malcolm,
I think they'll both be a lot closer than most people think, but I don't know if IR is the best place to compare them. The light setting are different (look at the highlights), focal length is different, jpeg engine is different(the comparator is using in camera jpegs) and the shutter speeds are different.

Of course what we are all really waiting for is the D800e test

Aaah... pixel peeping! I used the comparator to compare D800 to 5DMKIII at 1600 (more likely for wedding setups), and the detail in the fabrics is way better for D800. The Canon product does have less noise, but the Nikon noise is manageable as well.

In the end, does it matter? I don't see anything extraordinary for either brand, you know - the stuff to make people actually switch.

@malcolm - I would agree with you for the ISO=50 stills, but ISO=100 are arguably better on the D800. Just look at the fine texture of the fabric and the yarns. Admittedly, you have to pixel-peep to see it but what is the point of owning a 36MPx camera if you don't pixel peep (said he tongue in cheek)

I'm not sure what Malcolm is looking at, but in the comparator the Nikon looks better to me. But that's because the Canon JPG looks over processed. Too bad I don't have a raw processor for the Canon files yet. In RAW I imagine they both look fantastic.

The thing with the D800 that impresses me is how well it holds detail at high ISO even once some noise reduction is applied. I wasn't expecting results that good, despite have a D7000 with the same pixel pitch (that also impresses me).

Although I appreciate the effort, the problem I have with the tests is that a Sigma lens was used on a D800.

At 36mp, you'd better have a wicked high end Nikon lens to really make the tests worthwhile.

What I really want to see is a comparison between the D800 and Nokia 808 - including pixel peeping (said he with tongue in other cheek)

Yawn......Waiting for oly or fuji.....

Here we go again, my car is faster than yours, my dad is bigger than yours my Apple is better than your PC and on it goes. Seems to me that they are both superb tools. But the photographer makes the real difference! Please!

I was more interested in the differences between the D700 and D800. To me it's pretty clear that the D800 outresolves the D700. At ISO 6400 the D800 seems a little noisier at a per-pixel level, but it's still excellent and would probably look cleaner when resized to match the D700 at the same ISO.

All that detail and resolution, some pretty amazing stuff. Yet nothing so revolutionary that makes me want to upgrade (side-grade?) from my D700. Really, if I wanted high resolution, I would have bought the 5D Mark II with its 21MP when it first came out.

Whoa! Them's some mighty fine pixels right there.

Thank you Marten. Started to feel like I accidentally opened dpreview instead of TOP!

My D800E is on order. Of course, I had to place my order before good samples were available, so I was gambling a little. Nikon has put out some amazing cameras in recent years, and I was betting that they were continuing that trend. Looks like I am in luck.

I don't need 36MP. I don't even need 24MP. 18MP would have been my preferred figure, with more dynamic range and better high ISO performance, but beggars can't be choosers. In the end, I need more than the 12MP of my D3, due to occasional cropping, rotation, and so on, and so it was D4 vs. D800E. Due to the price difference I decided to try the D800E. If I don't like it, I will quickly resell it and buy a D3x.

I used to be a Canon owner, but to be honest, I don't understand Canon's choices and performance any longer. Why doesn't the same camera re-released years later (5DII, 5DIII) with minor updates blow the old one out of the water, no matter how you compare them? Has Canon hit a glass ceiling?

I've been holding on to my D200 since Dec 2005. While I've been using a GF-1 as my daily carry camera, the D800 has me itching to upgrade. I can easily see myself getting 6 years of use out of it, which makes the $3000 a bit easier to justify. The size of the RAW files does give me pause since I tend to keep every shot I take regardless of quality. Anyone else thinking of switching to a jpeg based workflow?

While I appreciate Shawn's reviews and in particular I-R's samples, I'm becoming more and more skeptical of these sorts of tests.

First there's the issue of operator technique - at these magnifications even minute differences in setup can make a huge difference. Then there's the problem of repeatability - most of us won't actually be using the camera in the lab with a macro lens on a tripod with mirror release. So I doubt we can realistically expect to see this sort of quality in our own results.

But yes, the are some mighty fine pixels...

Mike

Sounds like the Sigma 70mm they use for the test (which they use for all testing AFAIK) is a well-regarded lens and one of the sharpest SLRGear has ever reviewed. By all accounts it is as capable as any lens of delivering all the resolution the D800 is capable of capturing.

As a followup on my previous post I pulled down the Camera Raw 6.7 RC1. Opening the ISO 100 Canon 5D Mark III and D800 images with the same parameters you get very similar results as far as color. So my disagreement with Malcolm's interpretation is simply about the JPEG processing applied to the images. Both look fantastic, and very similar. The D800 does resolve a tiny bit more detail in this image. I will be interested to see real world results from both.

What really stood out to me in the sample files was the fine selection in beer! I am a poor hobbyist who is happy enough with his refurb'ed dSLR, so I am trying to avoid salivating over the latest-and-greatest. However, My hat's off to any appreciator of Samuel Smith. Bra-Vo, Imaging-Resource!

Looking at the Comparometer IMO my Canon 7D beats the 5D Mk II or III. The Mk II renders the fabric too soft for my taste and the Mk III too harsh. All three beat the D800. Again that's my taste in how I want images rendered.

I used the tool to do some pixel peeping of the D800 vs the 645D. While the 645D clearly out-resolves the D800, and with better contrast, the 645D is MF, goes for $10K, and doesn't have a low-pass filter.

A D800E vs 645D comparison will be interesting.

Anyone using the words "blows away" in a camera-related context should be made to sit in a dark gallery for 12 hours straight. 24 hours, without food, if the words "out of the water" are also invoked.

yeah, I am sure Nikon D800E with a proper lens will blow us away and out of the water into digital heaven right next to ... Nokia 808?

You know why the D800 looks worse at 100 iso: because it has 2 stops more D-range. You have to throw in a contrast curve to drop the upper and lower stop of light, and get an equal fresh looking picture. The killer feature of both the D7000 and the D800 is its +14 stop D-range. It is astonishing. You never got a nice sky like they should before on a DSLR. The D-range alone is worth the upgrade: bullocks to crap like high-iso, fps, AF systems etc.

Here is a nice selection from the linked images to illustrate it all:
http://imageshack.us/photo/my-images/826/5d3vsd800.jpg/
Left: 5D3, Right D800
The 5D has a black cork wrap with white lines, on the D800 they are what they should be: green with gold.
The 5D has too much sharpening, too much NR, too little D-range

The comments to this entry are closed.