if this has been posted, my apoloogies; if not enjoy the read!
http://www.dpreview.com/reviews/CanonEOS1DSMarkIII/

for the professionals that undoubtedly have to have the cutting edge tools, and hobbyists who are thinking "harley or camera for christmas," the 1DsMkIII is (almost) a no-brainer; but for the rest of the unwashed enthusiast-masses, one might want to consider the following numbers (from phil askey's DPR reviews of the 40D and 1DsMkIII):

40D
Features: 9.5
Image Quality: 9.0
Performance: 9.0
Price: $960

1DsMkIII
Features: 9.5
Image Quality: 9.5
Performance: 9.0
Price: $8000

Thanks, I didn't know it was up.

The last paragraph of the conclusion is interesting, Phil 'knows something' that he's not letting on about FF offerings to come I suspect

David

Interesting read...

A few things I noticed so far:

Phil's testing shows 1Ds3 high ISO noise performance is comparable to the 5D. This is what I found out as well. Actually, it is also very comparable to the D3 for that matter (up to ISO1600). He also found out that the luminance noise at higher ISO is very film-like after the color noise has been removed. I agree on that. Also, the level of details at higher ISO's remains very high.

Phil: "Overall softness that means images need more sharpening than you might expect; we'd prefer an even lighter Low Pass Filter."

I guess it depends on the lenses used, but this is not what I am seeing. I see at least comparable sharpness to the 5D (when files are equalized) and possibly slightly better.

abam wrote:
for the professionals that undoubtedly have to have the cutting edge tools, and hobbyists who are thinking "harley or camera for christmas," the 1DsMkIII is (almost) a no-brainer; but for the rest of the unwashed enthusiast-masses, one might want to consider the following numbers (from phil askey's DPR reviews of the 40D and 1DsMkIII):

40D
Features: 9.5
Image Quality: 9.0
Performance: 9.0
Price: $960

1DsMkIII
Features: 9.5
Image Quality: 9.5
Performance: 9.0
Price: $8000




Interesting......But you have to take into account that when giving the ratings, I'm sure they are making them in comparison to similar models ( at least in the features category).

Some point and shoot cameras also have the same ~9 point ratings, but this is not a proper comparison. JonStafford is right, this should be used to compare among similarly classed cameras.

Good review though. I've been wondering when it would come.

that could very well be, jon. i'm not sure if mr. askey has an explanation of his rating scale on DPR (or of the amount of subjectivity that goes into the final scores), but i'll see if a cursory glance through the site will yield anything.

cheers

http://www.dpreview.com/reviews/CanonEOS1DSMarkIII/page19.asp
The 5D and D3 results look a lot better than the 1Ds II !

Later on when shooting RAW the 1Ds II looks to being as well as the 1Ds III.

i stand corrected:

"Review ratings are given at the time of review (indicated beside the camera name) as an overall rating for the camera as it performed in our review tests, in comparison to other cameras in its class, price range and which were available / announced at the time of review. Ratings are the opinion of the reviewer. You must take the time to read reviews fully to be able to draw your own conclusion about a particular camera."

Pixel Perfect wrote:

http://www.dpreview.com/reviews/CanonEOS1DSMarkIII/page19.asp
The 5D and D3 results look a lot better than the 1Ds II !

Later on when shooting RAW the 1Ds II looks to being as well as the 1Ds III.

Yes, just soft in jpeg, but who shoot jpeg with 1DsII / III ?

KiboOst wrote:
Pixel Perfect wrote:

http://www.dpreview.com/reviews/CanonEOS1DSMarkIII/page19.asp
The 5D and D3 results look a lot better than the 1Ds II !

Later on when shooting RAW the 1Ds II looks to being as well as the 1Ds III.

Yes, just soft in jpeg, but who shoot jpeg with 1DsII / III ?

I do

The 1Ds3 JPEG quality and in-camera processing are very good. When the situation calls for it, I have no problem using JPEG.

Yes Daan, but they you set it to your taste, regardgin sharpening, saturation, hue etc, right ? You don't just let all in-jpegs settings to defaults ?
Apart dynamic range and 'pure' state of raw, that's why I never shoot jpeg : I rely on the body only for AF, iso, av, tv settings.

I don't see any interest in testing a camera in default jpeg mode. It is showing the standard settings from the bodie, not what we can do with the sensor capture.

KiboOst wrote:
Yes Daan, but they you set it to your taste, regardgin sharpening, saturation, hue etc, right ? You don't just let all in-jpegs settings to defaults ?
Apart dynamic range and 'pure' state of raw, that's why I never shoot jpeg : I rely on the body only for AF, iso, av, tv settings.

I don't see any interest in testing a camera in default jpeg mode. It is showing the standard settings from the bodie, not what we can do with the sensor capture.

Nope, I never use the defaults. I create my own user JPEG profiles in-camera. Although certain presets are not bad by themselves. The point I am trying to make is that the quality of JPEG's is the highest I have seen so far in a DSLR. Very usable. So, besides testing its RAW capabilities, I see no harm in testing how the JPEG engine performs. After all, it is a feature of the camera and some will actually use it

I thought the 5D's RAW looked better than the D3's! Oh I am not being biased...

Wow - the great Phil hasn't worked out that he needs a sharper lens to see a resolution difference between the 1Ds3 and 1Ds2:
"switching to raw and using the same processing (and crucially the same sharpening) shows that the resolution gap between the Mark III and the Mark II is minimal, and (aside from differences in depth of field) the output from the new camera is pretty similar, just bigger..."

Many of his 1Ds3 crops show the lens limiting sharpness.

To be honest, these days I treat DPReview with extreme suspicion. It contains so many technical errors.

JPEGs are very useful for a lot of applications. In the case of the 1Ds3, they are even more necessary if you shoot a lot, because raw files are quite slow to process even on top equipment.

I thought the D3 files looked rather soft on the test chart examples when compared to any of the other cameras. In camera NR anyone?

dhphoto wrote:

The last paragraph of the conclusion is interesting, Phil 'knows something' that he's not letting on about FF offerings to come I suspect

David

I agree. I believe he is getting it out there now, so it will be the basis to judge the new crop of cameras against. I also think that he wanted to give it the king of the hill award right now which it has deserved in case it gets dethroned.

brainiac wrote:
Wow - the great Phil hasn't worked out that he needs a sharper lens to see a resolution difference between the 1Ds3 and 1Ds2:


The review is done by Simon Joinson

brainiac wrote:

Many of his 1Ds3 crops show the lens limiting sharpness.



I agree totally.

In fact, his sample images at the end of the review show that the 1Ds MIII is not very fond of zooms. Even L zooms. At least the ones he used (or his particular copies of such lenses). Most of the sample images from the 24-70L and the 70-200L 2.8IS are somewhat on the soft side.
The ones taken with primes are definitely better, specially the small girl's portrait(5th pic from the right), and the beautiful black woman (7th pic from the left), where he used a 85mm 1.8.

While that is to be expected and considered normal, I think that this particular body does broaden the gap between primes and zooms in terms of IQ (sharpness wise, at least), more so than previous bodies.

As may have stated before, the 1DsMIII appears to demand the best optics for the best results.

fraga wrote:
As may have stated before, the 1DsMIII appears to demand the best optics for the best results.

Yes, whenever I read someone saying that the 1Ds3 requires extra sharpening, as Simon 'Philonikon' Joinson ('Phil' for short) does, I suspect they are seeing the limits of their lenses. You can't expect to get 21 megapixels of data through every lens, even at good apertures. If you want to see it pull away from the competition, you may need to avoid a lot of L zooms, and even some primes. It doesn't make the lenses worse than before, it just exposes their limits, at which point you're reviewing the lens, not the camera, as Simon Phil seems to spend much of his time doing.

Actually I think this review is one of the more Canon praiseworthy efforts, perhaps trying to redress the oft evident Nikon bias.

But then again, I'm biased too

David

abam wrote:
for the professionals that undoubtedly have to have the cutting edge tools, and hobbyists who are thinking "harley or camera for christmas," the 1DsMkIII is (almost) a no-brainer; but for the rest of the unwashed enthusiast-masses, one might want to consider the following numbers (from phil askey's DPR reviews of the 40D and 1DsMkIII):

40D
Features: 9.5
Image Quality: 9.0
Performance: 9.0
Price: $960

1DsMkIII
Features: 9.5
Image Quality: 9.5
Performance: 9.0
Price: $8000



the 1dsmkiii is certainly not a value buy, that said, i think they rate more within catergory I certianly would not put the 1dsmkiii IQ only .5 ahead of the 40D!

"...i think they rate more within catergory..."

yeah, i noticed that a bit too late. not that i expected the 40D to be on par with the 1DsMkIII, just found it interesting that one of the best-known reviewers had 'apparently' rated them fairly similarly. (...and as it ends up, apparently he did not.)

me=silly.

Pixel Perfect wrote:

http://www.dpreview.com/reviews/CanonEOS1DSMarkIII/page19.asp
The 5D and D3 results look a lot better than the 1Ds II !

clearly so but something is very wrong. A previous 5D/1Ds2 comparison by Phil using the same postage stamp had the 1Ds2 doing much better than it does in this test. see http://www.dpreview.com/reviews/CanonEOS5D/page21.asp

Also, if you look not at a sideways comparison between cameras but at a vertical comparison between ISOs for each camera then you'll see that the 5D loses more of its low-ISO detail as the ISO increases to 800 and beyond. To me a lot of its high-ISO clean image status seems to come at the expense of fine image detail.

I think too that the Standard picture style in the 1Ds3 and 5D is a bit punchier (more contrasty) than the default 1Ds2 non-picture-style settings or even the Neutral picture style. I'd prefer to see the comparisons done at comparable settings rather than default settings.

Later on when shooting RAW the 1Ds II looks to being as well as the 1Ds III.

which would indicate that the 1Ds2 jpeg softness is not a hardware issue because the raw images hold the details that the jpegs lost. They would not get the details if the AA filter was overly strong. It also implies that much of the apparent 1Ds3 (and 1D3) improvement is in the in-camera processing rather than the fundamental sensor technology, and that in turn leads to reasonable deduction that a 12Mpx 1Ds3 (or full frame 1D4) would be very clean at higher ISOs like the Nikon.

- Alan

Alan321 wrote:
>and that in turn leads to reasonable deduction that a 12Mpx 1Ds3 (or full frame 1D4) would be very clean at higher ISOs like the Nikon.

The 1Ds3 is very clean at 12Mpx, like the Nikon. Just shoot at 21Mpx and downrez to taste. Be sure to do any chroma noise reduction before downrezzing.

Here are DPReview's iso 3200 crops at equal magnification, 1Ds3, 1Ds3, D3: