michaelGIORDANO wrote:
It looks to be true...

Nikon's D800 outclasses camera rivals in sensor test. The new high-end SLR sets a top overall score in DxO Labs' image sensor tests, beating even medium-format cameras.

http://news.cnet.com/nikons-d800-outcla … 5-264.html

Nice. Did you actually go to DxO?  You'll find they haven't tested the 5D3 yet nor when this article was written.

michaelGIORDANO wrote:
It looks to be true...

Nikon's D800 outclasses camera rivals in sensor test. The new high-end SLR sets a top overall score in DxO Labs' image sensor tests, beating even medium-format cameras.

http://news.cnet.com/nikons-d800-outcla … 5-264.html

sublime LightWorks wrote:
Nice. Did you actually go to DxO?  You'll find they haven't tested the 5D3 yet nor when this article was written.

Your point is valid.  As of now, the D800 outclasses everything that has been tested so far.  We'll have to wait and see.  You are correct, the 5D III could be better or not!  Only time will tell.

ei Total Productions wrote:
Your point is valid.  As of now, the D800 outclasses everything that has been tested so far.  We'll have to wait and see.  You are correct, the 5D III could be better or not!  Only time will tell.

True, but evidence so far suggests that the 5D III's sensor will score lower than the D800's in most tests.

KFM Designs wrote:
http://www.youtube.com/watch?v=omTo7UxbJX8

Pretty cool

Part 2:
http://www.youtube.com/watch?v=4W9EeDCaVFM

K E S L E R wrote:
I don't think this comparison or 99% of the comparisons out there between the 5D3 vs D800 even cares about real world use.

Its about sensor.  A sensor 66% larger than what its being compared to...  But seriously, I'd love to see this sensor comparison between the D4 vs the 5D3.

Both the 5D & D700 were compact FF D-SLR's. The D4 is Nikon's flagship pro-body D-SLR. Why compare a compact to a pro-body??? (Note, I'm referring to where they sit in the lineup's, and what they were designed for - NOT their use by pro's in the real world!)

IMHO alone, as always;

~Danny
http://www.dbiphotography.com/
https://www.modelmayhem.com/2401686

Where the D700.5D2 & D800/5D3 sit in their lineups: http://gizmodo.com/5160540/canon-5d-mar … -shoot+out

D3
1/80th
55mm focal length
f/4.5
ISO 4000

Leggy Mountbatten wrote:
As someone who owns both the MK II and the MK III, I can tell you definitively that high ISO performance in RAW is significantly better in the Mk III. I rate a camera's ISO performance by its "speed limit." In other words, what's the maximum ISO that would be useful to me?

The Mk II has a hard limit of 6400, because pattern noise becomes dominant above that ISO. It's the same limit as the 7D; above that, and the images turn to complete garbage.

For me, the Mk III's entire ISO range is now useful, depending on circumstances. That's a four stop difference. That's huge.

Before someone jumps in and claims I'm saying that ISO 102k on the Mk III is the same as 6400 on the Mk II: that's not what I'm saying. Just that the pattern noise is pretty much gone at the standard ISO's, and something I can deal with in Lightroom at ISO 51k and 102K. ISO 25K isn't even remotely usable in the Mk II; in the Mk III, it's quite respectable. I'd have no issues selling images shot at any ISO in the full range of the camera.

I don't have mark III but I do have mark II.  From all the info I have seen, the Mark III's strength is only JPG.  Canon cooked JPG file well.  My guess is Canon's 5DIII's sensor will not even get higher than 13 DR.  It probably be around 12.5 or less.  But that is just guessing. 

  Besides, if you looking for Camera with high ISO performance, shouldn't you be looking into 1DX?   Why even bother with the 5DIII when the goal is for high ISO performance.

ChanStudio - OtherSide wrote:
I don't have mark III but I do have mark II.  From all the info I have seen, the Mark III's strength is only JPG.  Canon cooked JPG file well.

I only shoot RAW. Only. Just RAW. No jpeg's. Just wanted to be clear with you. The difference I'm seeing is when processing the files in Adobe Lightroom, processing the RAW files.

ChanStudio - OtherSide wrote:
My guess is Canon's 5DIII's sensor will not even get higher than 13 DR.  It probably be around 12.5 or less.  But that is just guessing.

The ISO 100 files have a bit less pattern noise in the shadows than the Mk II, but otherwise, the performance seems pretty similar. But, then, I'm shooting at ISO 800 and up the majority of the time, and that's where I'm seeing very significant improvements in the RAW files.

ChanStudio - OtherSide wrote:
Besides, if you looking for Camera with high ISO performance, shouldn't you be looking into 1DX?   Why even bother with the 5DIII when the goal is for high ISO performance.

Well, let's see...

1. The 5D Mk III is available today. The 1Dx won't be available until June.

2. The 1Dx is twice the price.

3. There's no evidence, so far, that the 1Dx is superior to the 5D Mk III in low light performance. Perhaps it will be, but thus far, the evidence is inconclusive.

4. Who says I'm not going to get a 1Dx once they're available?

5. The performance of the Mk III at stratospheric ISO's is remarkably good; the improvement in high ISO performance for the 1Dx, if there is any, may not be worth the cost.

Here's an example shot at ISO 102400. This was not a Canon jpeg, instead being a RAW that I processed in Adobe Lightroom:



Did I mention that I only shoot RAW?

Leggy Mountbatten wrote:
http://galleries.stevemelvin.com/5d_test/

Not an ideal test, but there are some things that can be gleaned from it.

Compared all three generations of 5D cameras, with links to images all resized to Mk I native size (4368 pixels wide).

ISO's from 200 to 102000 from all three cameras. Higher ISO's created by underexposing and pushing in Lightroom for the Mk I and Mk II.

Also did an image at ISO 100 underexposed by three stops and pushed in Lightroom to try to make it look like it was correctly exposed. It appears that the Mk II handles the push worse than the Mk I or Mk III, though none of them really passes this test. This is really the one thing I would like Canon to fix.

The Mk III seems to recover just a little more highlight detail than the others. On all cameras, the highlight recovery at ISO 100 and 125 was identical, and 160 was able to recover less detail. But 160 has more shadow detail. This lends support to the theory that 160 is a pull from 200, and 125 is a push from 100.

All shot with Auto WB, to see if they behaved differently. The Mk I and Mk II behaved identically, giving a color temp of 3550 and tint of +7. The Mk III gave a color temperature of 3250 and tint of +12. Perhaps the new meter is playing a role in this.

I used identical shutter speed and apertures for all three cameras; exposures appear to be identical over the standard ISO range.

The Mk III is clearly giving a lot more shadow detail at higher ISO's, not to mention much more pleasing noise. Note the cinder blocks to the left of the chair at ISO 100k.

All cameras are showing "amp glow" in the corners at extreme ISO's. Fortunately, the corners are usually relatively easy to deal with.

I tested, but didn't post, shots at 100, 125 and 160. I used a Fill Light of 80 to look at the shadow noise. On the Mk I and Mk II, 125 has the worst-looking shadow noise. 160 tends to look best, but not a whole lot better than 100. On the Mk III, the contrast changes, with 160 having the lowest contrast and most shadow detail. Again, 125 looks the worst.

The Mk I gives less-than-expected brightness at ISO 3200. I adjusted the ISO 6400 (equivalent) exposure to match the brightness of the ISO 1600 exposure.

Thank you so much for doing these and I'm sorry for missing them first time round

I agree with your assessment - there's more DR (cleaner shadows) in the higher ISO shots on the 5d3. Underneath the chair here is a really good example:

http://galleries.stevemelvin.com/5d_test/Mk-2-6400.jpg
http://galleries.stevemelvin.com/5d_test/Mk-3-6400.jpg

It's fine having a shot with nice highlights at 6400, but if the shadows look poor it's virtually unusable. That's great news then!

The bigger difference though is at 12800:
http://galleries.stevemelvin.com/5d_test/Mk-2-12800.jpg
http://galleries.stevemelvin.com/5d_test/Mk-3-12800.jpg

One is usable and the other pretty much isn't. The mk2 has just mostly fallen apart with poor quality noise and all sorts of complete misreads of the signal but again, DR on the mk3 looks "reasonable" considering what it's shot at. It's a usable file.

3200 showed the same - the mid to deep shadows had more detail in the mk3 file. The wood behind the vertical "gate" on the bottom right shows this well. Great news for wedding photographers who regularly shoot at 3200 - retaining as much real detail in the shadows as possible is key since skin tones will be at these brightness levels at times.

Detail under the chair at 800 looked better on the mk3, but it was a struggle to tell them apart.

Yes - low ISO DR is what Canon clearly hasn't put their R&D efforts into... but what I'm expecting to see in the dxo etc.. tests is that DR at all ISOs past 800 have been improved because I can see it. Again that's great news for anyone who shoots at those ISOs regularly.

This camera really does seem to be targeted towards event photographers: class leading AF, improved DR at high ISO.. I think Canon realised this was the section of the market which was particularly struggling with the 5d2 (although I still maintain the centre point of the 5d2 was fine until things got very dark)

ChanStudio - OtherSide wrote:

I don't have mark III but I do have mark II.  From all the info I have seen, the Mark III's strength is only JPG.  Canon cooked JPG file well.  My guess is Canon's 5DIII's sensor will not even get higher than 13 DR.  It probably be around 12.5 or less.  But that is just guessing. 

  Besides, if you looking for Camera with high ISO performance, shouldn't you be looking into 1DX?   Why even bother with the 5DIII when the goal is for high ISO performance.

I think it's true that the MKIII images don't look much different from the MKII images. The thing is, they look just like the 5DII's images two ISO stops down, or more.

The equivalent high ISO images from the 5DII don't exist because the AF is inadequate to take them.

I've got some amazing images from the 5DII shot at 12,800. They look like a 7D around 1600, but they were shot in daylight.

High ISO performance and low light performance are not the same thing.

Odin Photo wrote:
I skipped to the end, so I apologize if this has already been covered, but that is the same image that has been processed. Has no one else noticed that the noise is the same in both images?

No, because it isn't.

Wherever you got that gif is full of it.

The images come from dpreview. You can download the raws and re-create the gif for yourself. It's possible the OP made a scaling mistake; would be nice if someone clarified.

Odin Photo wrote:
I skipped to the end, so I apologize if this has already been covered, but that is the same image that has been processed. Has no one else noticed that the noise is the same in both images? Look at the biggest noise pixels, and guess what? They are in the same positions on the D800 inset. Please don't tell me this is a coincidence.

Wherever you got that gif is full of it. It is a fake OR someone is blending the layers instead of using 100% opacity in the layers making the gif.

The fact that the OP is suggesting that the D800 is better at ISO 12,800 than the 5DIII reveals what little and limited knowledge he has despite his assertions.  I have not seen a single reputable review (or any review for that matter) or comparison between the two that would suggest that the D800 is better at ISO 12,800.  This should be a red flag for anyone reading the OP's "analysis" of this or any equipment in the future.

Next we will be hearing that the D800 has faster FPS... lol..

They are both excellent camera's that offer their own advantages.  Do your homework by educating yourself with reputable reviewers and take absurd claims such as this, as nothing more than an attempt to garner attention.

Moon Pix Photography wrote:
Do your homework by educating yourself with reputable reviewers..........

In the same vein, it would do some well to research posters who are posting before referring to them in a derogatory manner.

~Danny
http://www.dbiphotography.com/
https://www.modelmayhem.com/2401686

DBIphotography Toronto wrote:

Both the 5D & D700 were compact FF D-SLR's. The D4 is Nikon's flagship pro-body D-SLR. Why compare a compact to a pro-body??? (Note, I'm referring to where they sit in the lineup's, and what they were designed for - NOT their use by pro's in the real world!)

IMHO alone, as always;

~Danny
http://www.dbiphotography.com/
https://www.modelmayhem.com/2401686

Where the D700.5D2 & D800/5D3 sit in their lineups: http://gizmodo.com/5160540/canon-5d-mar … -shoot+out

My guess is because these ratings are rating the SENSOR and not anything else.  In which case, the body it is attached to becomes irrelevant as to its features or usage.

Leggy Mountbatten wrote:
Very well-done real world comparison.
http://www.youtube.com/watch?v=4W9EeDCaVFM

That was awesome!

Sendu wrote:

Odin Photo wrote:
I skipped to the end, so I apologize if this has already been covered, but that is the same image that has been processed. Has no one else noticed that the noise is the same in both images?

No, because it isn't.


The images come from dpreview. You can download the raws and re-create the gif for yourself. It's possible the OP made a scaling mistake; would be nice if someone clarified.

Yes, it is. In the gif the >patterns< of noise are the same. The noise is diminished in the inset, but the prominent pixels and pixel clusters of noise are very easy to see. It is like the image was run through a noise cleaning program and had sharpening run on it but there are portions of noise that because of their size and color contrast are very obviously in the exact same positions even after the noise has been diminished. Either dp review (yes I am very familiar with the website and it's testing practices) messed up or the information or technique used to create the gif was messed up.

This would absolutely not happen with images from two different cameras, and even more so from two different brand of cameras. The noise patterns remaining through the process to make the gif is practically a finger print saying "Hey Folks, same source image here" or perhaps, "Blended layers image here". Either way it is NOT a relevant comparison between two different image sources.

Odin Photo wrote:
Either dp review (yes I am very familiar with the website and it's testing practices) messed up or the information or technique used to create the gif was messed up.

Why speculate and make poorly founded accusations and claims, when you can just prove it one or the other by looking at the original files yourself?

KFM Designs wrote:

That was awesome!

There's a part one too http://youtu.be/omTo7UxbJX8

I think both articles have come to the same conclusion I have - they're both AMAZING cameras and the differences between them are smaller than ever before.

Sendu wrote:
Why speculate and make poorly founded accusations and claims, when you can just prove it one or the other by looking at the original files yourself?

Why come here and suggest someone else do all the work, instead of simply proving me wrong by looking at the original files yourself? It doesn't seem like you want to follow your own advice, so I did the work for you.

Before we get to the work part, it's not a speculation, it is an extrapolation, well based on the obvious information in the gif provided. The gif itself has enough information to prove it is using the same image or at least having a portion of transparency showing through because of the way it was constructed or saved, which renders it useless as a comparison image anyway. Even after converting to a gif, they wouldn't have the same noise artifacts in the same locations. Don't pay attention to the watch features. Look specifically at the color/luminosity noise  pixels in the gif and it is obvious. If you don't see it, then you can investigate, just like I did below.

I just went to DPReview and downloaded both files from the D800 and the 5D MKIII and super imposed them after scaling the D800 down to match the original 5DMKIII and also scaling the 5DMKIII up to match the original D800. Using both techniques, neither looks like the gif here. 

Also if you download the files from DPReview you will see that the color checker card is in a different relative position for the different cameras than represented in the gif. Not by a great amount, but enough to be obvious. At this point, it should be quite clear that this gif is not a representation of the two images that can be downloaded from DPReview for the D800 at iso 12800
movies.dpreview.com.s3.amazonaws.com/nikon_d800/DSC_0261.NEF.zip
and the 5DMKIII at 12800
movies.dpreview.com.s3.amazonaws.com/canon_eos5dmkii/IMG_0059.CR2.zip
Just like I knew from the beginning. 
You're welcome.

Odin Photo wrote:
Why come here and suggest someone else do all the work, instead of simply proving me wrong by looking at the original files yourself?

I wasn't making any claim. I was asking you to support yours.

I just went to DPReview and downloaded both files from the D800 and the 5D MKIII and super imposed them after scaling the D800 down to match the original 5DMKIII and also scaling the 5DMKIII up to match the original D800. Using both techniques, neither looks like the gif here.

So you did the work, but don't show us the result? So you still haven't supported your claim, instead insisting we just take you at your word. That strategy still isn't working.

Sendu wrote:

Odin Photo wrote:
Why come here and suggest someone else do all the work, instead of simply proving me wrong by looking at the original files yourself?

I wasn't making any claim. I was asking you to support yours.


So you did the work, but don't show us the result? So you still haven't supported your claim, instead insisting we just take you at your word. That strategy still isn't working.

If I am going to lie to you through the text on this page, why wouldn't I lie to you with a gif of my own making?

I gave you all the information you need to see that you are wrong for yourself, rather than taking my word (or gif) for it. I have supported my claim perfectly by removing myself from the equation and guiding you directly to the source files without having an opportunity to alter them. All that is required of you is a small effort following some links or comprehending what I said earlier and looking at the gif with a critical eye.

ei Total Productions wrote:
Your point is valid.  As of now, the D800 outclasses everything that has been tested so far.  We'll have to wait and see.  You are correct, the 5D III could be better or not!  Only time will tell.

They are planning the 5DmIII test. I did see in other tests for video that the mIII held up better in low light. (motion not stills)  I think it can be subjective. Depends on how we light things.  I like how this looked (shot exclusively on the Nikon D800): http://vimeo.com/36326055

Here is the behind the scenes footage:  http://vimeo.com/36327457

Robb Mann wrote:
This thread is still alive? Wow.

Crazy HUH.......

michaelGIORDANO wrote:
Results are in. They finally tested the latest Canon and the conclusion is the 5DMIII is the best among all Canon's but still falls short of the Nikon D800.

http://www.petapixel.com/2012/04/20/dxo … taPixel%29

I think you have just made a very fair and objective statement.  The results also vindicate Fred's contention.

ei Total Productions wrote:

I think you have just made a very fair and objective statement.  The results also vindicate Fred's contention.

It all depends how you plan to use the camera.  Each one has different strengths.
One may work better for you than the other.

ei Total Productions wrote:
I think you have just made a very fair and objective statement.  The results also vindicate Fred's contention.

Jerry Nemeth wrote:
It all depends how you plan to use the camera.  Each one has different strengths.
One may work better for you than the other.

I thnk he was just referring to DXOmark sensor rankings.  I don't think that he, or I, were talking about which camera is better.  I have already said that I think that the 5D III is probably better for sports and the D800 better for studio.   I don't think that any camera is perfect.

There is a lot more to a camera than the DXOmark sensor rating.  I think we all know that both Canon and Nikon make great camras.

Your point is well taken.  It needs to be mentioned that the comments import is limited to just that subject.

ei Total Productions wrote:

ei Total Productions wrote:
I think you have just made a very fair and objective statement.  The results also vindicate Fred's contention.

I thnk he was just referring to DXOmark sensor rankings.  I don't think that he, or I, were talking about which camera is better.  I have already said that I think that the 5D III is probably better for sports and the D800 better for studio.   I don't think that any camera is perfect.

There is a lot more to a camera than the DXOmark sensor rating.  I think we all know that both Canon and Nikon make great camras.

Your point is well taken.  It needs to be mentioned that the comments import is limited to just that subject.

Focusing on only the DXOmark sensor ranking is like looking at the cameras with tunnel vision.

1ST CLASS  Photography wrote:
I saw an ISO comparison wear the D800 was equal with the D4 when the images were viewed at the same size on the monitor, and i saw another comparison wear they were saying the image of the D800 is more amazing than the D4, and I saw a video comparison at 12800 iso wear the canon 5d3 blew away the D800 as far as noise, does anyone know wear you can see image comparisons, not just ISO comparisons, but detail and depth and the overall beauty of the image ?  I am not a fanboy of either brand, I love both cameras, I know they both have there own ups and downs, but I am trying to figure out witch one has the most ups  ? 
also I agree with whoever said it would be good to see the 4D compared with the 5D3

Personally I would tend to look at the dynamic range at each ISO as part of the answer. DXO does a reasonable job of this (although other sites disagree on the exact results).

At high ISO pecking order would be something like:
D3s > D4 > 5d3 > D800 > D700

It doesn't take everything into account (such as colour accuracy) but it'd give you a good feel for the results. There's only a stop between pretty much all of the cameras at all ISOs over 800.

In terms of downloading files, I believe you can do that at http://www.imaging-resource.com/ . That might give you what you need?