Ad
Ad
Ad
Pages: « 1 [2] 3 4 ... 6 »   Bottom of Page
Print
Author Topic: Canon Mk3,P21 and ZD compared  (Read 24388 times)
samuel_js
Guest
« Reply #20 on: December 03, 2007, 04:33:26 AM »
ReplyReply

Deleted
« Last Edit: December 05, 2007, 12:57:58 PM by samuel_js » Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8900


« Reply #21 on: December 03, 2007, 07:27:51 AM »
ReplyReply

Quote
Could you guys forget the numbers for a moment and look at the pictures? Of course the speeds and apertures are not optimal, but all three losses from that, not only the canon. My H2 at 5,6 is much sharper than at 2,8... But this is what we have....

[a href=\"index.php?act=findpost&pid=157848\"][{POST_SNAPBACK}][/a]

Forget the numbers?? Forget the fact we're comparing a 17.23mp sensor of almost 2 1/2 times the size with a 14.47mp sensor, after cropping to the same FoV? Forget the fact that the 1Ds3 appears to be focussed at infinity whereas the P21 appears to be focussed at some hyperfocal distance producing a sharper foreground?

If you want to search for an area in the image where the P21 produces a very marginally sharper result on huge size prints, you can do better than the crops you've shown.

How about the following 2 crops where even the softer ZD looks sharper than the 1Ds3?

[attachment=4112:attachment]  [attachment=4113:attachment]

In the first example above I tried applying a bit of sharpening. 150% at 1.5 pixels for the ZD, 100% for the P21 and 75% for the 1Ds3.

The second example is unsharpened as are the crops below which show identical detail in the P21 and 1Ds3 image and a noticeably softer result in the ZD.

[attachment=4114:attachment]

The 1Ds3 is not inferior in any way to the P21 in these tests considering how it's been handicapped by a flawed procedure.
Logged
samuel_js
Guest
« Reply #22 on: December 03, 2007, 07:38:29 AM »
ReplyReply

Deleted
« Last Edit: December 05, 2007, 12:58:22 PM by samuel_js » Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8900


« Reply #23 on: December 03, 2007, 08:01:22 AM »
ReplyReply

Quote
Do you own a DB? I suppose this thread is gonna end like the famous DB vs 35 mm, "the 1ds is as good as the the digital backs", but in fact, most of those defendors has never tried one.
[a href=\"index.php?act=findpost&pid=157873\"][{POST_SNAPBACK}][/a]

Samuel,
I can't following your reasoning here. I haven't tried 99% of the various camera models, brands and formats on the market, as probably most readers on this forum haven't.

Does that mean I'm not capable of comparing the quality of 2 images side by side?

The only thing I'm defending here is my sense of objectivity and impartiality.

It is a fact that there are areas of these images where the P21 is marginally sharper and more detailed than the 1Ds3 image and that applies also to the ZD image although not to the same extent. If this wasn't the case, it would be a bit strange considering the P21 has a greater pixel count in relation to the same FoV, has a lens of probable equal quality and therefore should produce images of greater accutance since the larger sensor is accessing a lower resolution at a higher MTF, and significantly does not have an AA filter.

Merely the lack of an AA filter in the P21 could account for some of that marginally greater accutance on signs in the foreground.

It's not certain if different focussing is contributing to the resolution differences. I mention it as a possibility. However, it is a fact that a 1Ds3 at f11 will have greater DoF and even more so considering the the FoV of the 1Ds3 shot is slightly wider in these comparisons in both dimensions but particularly in length.
« Last Edit: December 03, 2007, 08:39:12 AM by Ray » Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8900


« Reply #24 on: December 03, 2007, 09:01:02 AM »
ReplyReply

Just as a matter of interest, because the numbers must reflect on these comparisons, if the FoV in the short dimension had been matched between the P21 and the 1Ds3, we'd be comparing a 17.23mp P21 image with a 17.8mp 1Ds3 image. Such a comparison would favour the P21 because the 1Ds3 is a 21mp camera, but would still be fairer than the current comparison which matches a 17.23mp P21 image with a 14.47mp 1Ds3 image.

If we match the FoV along the longer dimension of each camera, we would in effect be comparing a 15.36mp P21 image with a 21mp 1Ds3 image, which of course gives the advantage to the 1Ds3.

We need to know how these cameras compare in these ways before any definite statement can be made about image quality.
Logged
routlaw
Jr. Member
**
Offline Offline

Posts: 88


WWW
« Reply #25 on: December 03, 2007, 10:12:08 AM »
ReplyReply

Thought you folks might be interested in looking at the blue channel in these two camera files given that these two seemed to be creating quite the heated debate.

Rob
Logged
routlaw
Jr. Member
**
Offline Offline

Posts: 88


WWW
« Reply #26 on: December 03, 2007, 10:15:54 AM »
ReplyReply

Thought you folks might be interested in looking at the blue channel in these two camera files given that these two seemed to be creating quite the heated debate.

Rob[attachment=4115:attachment]
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8900


« Reply #27 on: December 03, 2007, 10:32:36 AM »
ReplyReply

Quote
WHO CARES!

[a href=\"index.php?act=findpost&pid=157910\"][{POST_SNAPBACK}][/a]

I guess all those who want the best tool for the job care as well as all those who are careful with their spending. The fact that differences may be hair splitting is useful to know, but you can't know that without first splitting the hairs.

The differences between the 1Ds MkII and the 5D were hairsplitting and people needed to know that in order to make an informed purchasing decision. Some folks were prepared to spend the extra money for the other advantages of the 1Ds2 such as greater ruggedness, better focussing, waterproofing etc. Others seemed to prefer the 5D, despite the fact it was a much cheaper camera, because it produced more pleasing skin tones.

My own view is, when comparing moderately different formats with the same pixel count, the quality of the lens is going to make the greatest difference.

The comparisons at the head of this thread are hoplessly biased and people need to be aware of this. One advantage of the smaller format is its capacity to deliver equivalent DoF at a wider aperture. In addition to the mismatch of FoVs in this comparison, the 1Ds3 should have been using f8 instead of f11 in order to match DoFs.
Logged
samuel_js
Guest
« Reply #28 on: December 03, 2007, 11:17:21 AM »
ReplyReply

Deleted
« Last Edit: December 05, 2007, 12:58:56 PM by samuel_js » Logged
203
Guest
« Reply #29 on: December 03, 2007, 11:22:02 AM »
ReplyReply

Quote
WHO CARES!

 That's fantastic. What's not great is the endless and pointless splitting hairs on resolution. Which is just only a part of the whole equation to any digital capture device.

[a href=\"index.php?act=findpost&pid=157910\"][{POST_SNAPBACK}][/a]


Indeed. Good luck telling the difference between these cameras in print, fellows.

This is clearly a self-selected group of gear heads here...

By the way John, have you tried the Chromatic Aberration and other new tools in the latest versions of DPP? If so, do they not help with your 85 1.2 issue?

-best
« Last Edit: December 03, 2007, 11:23:08 AM by 203 » Logged
routlaw
Jr. Member
**
Offline Offline

Posts: 88


WWW
« Reply #30 on: December 03, 2007, 11:31:36 AM »
ReplyReply

But look at the area with grass, where the P21 is blocked up with noise vs the Canon. BTW I don't own either camera or for that matter cameras from either company. I do agree with others on the thread the comparison and test left a lot to be desired. I doubt any of these differences would show up when you put ink to paper, be it offset press or high end inkjet.

Rob

Quote
Again, the blue channel only reveals more detail in the P21.
Logged
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2427



WWW
« Reply #31 on: December 03, 2007, 11:54:58 AM »
ReplyReply

Because of the differences in lenses, and the differences in distance at which they were focused.  Same goes for the raw conversions and jpeg compressions.  

With all that in mind, I'll still venture to say that there are real differences in color and DR that can be observed and which some have already pointed out.  Whether this is significant depends only on you or your clients.
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
alba63
Jr. Member
**
Offline Offline

Posts: 66


« Reply #32 on: December 03, 2007, 12:00:59 PM »
ReplyReply

Things get much easier if people would just agree that in real life shooting the 1ds3 will come quite close to the MF backs - at least the backs up to 22MP. Everyone knows that the Canon has a smaller Sensor, an AA filter and will therefore have less sharpness, slightly less resolution, and less dynamic range.

However wondering how much of this difference translates in visible print quality, the real question will be how many shooters will prefer the Canon for much more speed + versatility, much lower ISO noise, faster AF, 10 times more lenses and a lower price etc.

Therefore "quite close" is probably all that Canon wanted it to be compared to the MF world.

it also seems a bit strange to compare the output of 3 cameras that cost 7k the cheapest, with the help of 2MB JPEGs?
The only one of the 3 I have shot was the ZD (SLR version) and looking at my few handheld quick shots they are yet far above anything I have shot with my 5d, in terms of sharpness and detail. Those were made with a cheap manual 55mm Mamiya lens.

Bernie
Logged
Snook
Guest
« Reply #33 on: December 03, 2007, 12:29:21 PM »
ReplyReply

Quote
Things get much easier if people would just agree that in real life shooting the 1ds3 will come quite close to the MF backs - at least the backs up to 22MP. Everyone knows that the Canon has a smaller Sensor, an AA filter and will therefore have less sharpness, slightly less resolution, and less dynamic range.

However wondering how much of this difference translates in visible print quality, the real question will be how many shooters will prefer the Canon for much more speed + versatility, much lower ISO noise, faster AF, 10 times more lenses and a lower price etc.

Therefore "quite close" is probably all that Canon wanted it to be compared to the MF world.

it also seems a bit strange to compare the output of 3 cameras that cost 7k the cheapest, with the help of 2MB JPEGs?
The only one of the 3 I have shot was the ZD (SLR version) and looking at my few handheld quick shots they are yet far above anything I have shot with my 5d, in terms of sharpness and detail. Those were made with a cheap manual 55mm Mamiya lens.

Bernie
[a href=\"index.php?act=findpost&pid=157957\"][{POST_SNAPBACK}][/a]
Doesn't Sensor size and camera format have influence in anything? I thought the Bigger sensor size and Lens are going to be better on MFDB for sure? Or am I missing something?
Snook
Logged
John Camp
Sr. Member
****
Offline Offline

Posts: 1259


« Reply #34 on: December 03, 2007, 12:36:12 PM »
ReplyReply

The attacks on the test conditions are silly -- there are no perfect tests, but that doesn't mean you can't reach a few conclusions. One conclusion I've reached is that none of the 35s, because of the trade offs between pixel count/pixel size/resolution/DR on one hand, and noise on the other, will ever be as absolutely good as the top MF backs with good lenses.

But so what? That's meaningless except for the tiny fraction of people who BOTH afford MF backs AND make their living from fine art prints. The evidence seems to be piling up that the end product for *most* commercial photography -- small- to medium-sized prints, magazine shots or web photos -- there is no discernible difference in the final product between shots with a camera like a 1DsIII and a P45.

There will, of course, be a visible difference in large, one-off, digital prints. But I would suggest that even that difference won't be particularly visible if you're doing photos for a store promotion (like Victoria's Secret, which uses poster-sized model shots in the windows) and the posters are printed on commercial printing presses of less than the highest fidelity.

There is, in fact, a practical aspect to all of this, and for those practical purposes, MF may have more and more difficulty proving that it is a desirable option vis-a-vis 35.

Of course, there are other aspects, like art directors who demand that the shots be done on MF, because they believe that they need the highest possible quality, even when the final product will be printed on toilet paper...

JC
Logged
rainer_v
Sr. Member
****
Offline Offline

Posts: 1131


WWW
« Reply #35 on: December 03, 2007, 12:52:43 PM »
ReplyReply

Quote
The attacks on the test conditions are silly -- there are no perfect tests, but that doesn't mean you can't reach a few conclusions. One conclusion I've reached is that none of the 35s, because of the trade offs between pixel count/pixel size/resolution/DR on one hand, and noise on the other, will ever be as absolutely good as the top MF backs with good lenses.

But so what? That's meaningless except for the tiny fraction of people who BOTH afford MF backs AND make their living from fine art prints. The evidence seems to be piling up that the end product for *most* commercial photography -- small- to medium-sized prints, magazine shots or web photos -- there is no discernible difference in the final product between shots with a camera like a 1DsIII and a P45.

There will, of course, be a visible difference in large, one-off, digital prints. But I would suggest that even that difference won't be particularly visible if you're doing photos for a store promotion (like Victoria's Secret, which uses poster-sized model shots in the windows) and the posters are printed on commercial printing presses of less than the highest fidelity.

There is, in fact, a practical aspect to all of this, and for those practical purposes, MF may have more and more difficulty proving that it is a desirable option vis-a-vis 35.

Of course, there are other aspects, like art directors who demand that the shots be done on MF, because they believe that they need the highest possible quality, even when the final product will be printed on toilet paper...

JC
[a href=\"index.php?act=findpost&pid=157963\"][{POST_SNAPBACK}][/a]

well said.
.... but maybe its 44" long fineart toilet paper from hahnemuehle for real rich clients ....
Logged

rainer viertlböck
architecture photographer
munich / germany

www.tangential.de
Ray
Sr. Member
****
Offline Offline

Posts: 8900


« Reply #36 on: December 03, 2007, 12:59:40 PM »
ReplyReply

Quote
The attacks on the test conditions are silly -- there are no perfect tests....[a href=\"index.php?act=findpost&pid=157963\"][{POST_SNAPBACK}][/a]

Whaddaya mean? All my tests are perfect.  

The fact is, there are certain factors one has no control over, such as manufacturing QC variations between cameras and backs of the same model, differences in lens quality when comparing different brands and formats of cameras with different lenses of different focal lengths, differences due to the presence or lack of an AA filter and so on.

Such factors are difficult to take into consideration, although I believe the ZD does have an optional AA filter which I suspect was not used in this test.

However, there are factors which are not difficult to take into consideration. If you can do it, then why not do it?

Matching FoV is a very basic thing to do. Throwing away almost 1/3rd of a camera's pixels in the process of making a comparison is not merely 'not perfect' but a major blunder.
Logged
samuel_js
Guest
« Reply #37 on: December 03, 2007, 01:06:59 PM »
ReplyReply

Deleted
« Last Edit: December 05, 2007, 12:59:32 PM by samuel_js » Logged
rainer_v
Sr. Member
****
Offline Offline

Posts: 1131


WWW
« Reply #38 on: December 03, 2007, 02:24:15 PM »
ReplyReply

i think the resolution increase in comparation to the 1dsmk2 is bigger than the pure pixel count had let me to think. this can only mean that the aa filter is better and much weaker designed. this was the point i was most sceptical with the mk3 design,- but it seems so that canon was aware that they calculated the aa filter  too strong in their 1ds mk2 model,- ( therefor the 5d could render the same detail with less megapixels ).

the DR seems also to be excellent from the new mk3 .
the most impressive detail in the comparation above is the look in the blue channel .......

it seems so that there is zero or very little difference between the usable dr comparing the canon with the actual mf backs at lowest isos,- maybe the canon even wins here, this will depend also if streaks or bandings may appear, which usally will take some days or weeks till we "beta buyers" will find them, if not the canon will not only match most mf backs it will be even be better than the "lemons" which are sold from the wide production tolerances.

color profiling is a thing which should be done individualy in any case, with mf backs and with 35mm cameras,- its hard to say how successfulll this can be made, but lightromm is a great konverter to realise every desired "look".
in my opinion there is much less room for the mf manufactors now,-
which is a pity for me, because for architecture or studio photography the mf advantage is less caused by the sensor as by the available lenses,- which are not to be expect to become available in 35mm systems for the system immanent retrofocal constructions. also the step from 22 to 33/39mp is not so little, at least if the sharpest lenses are used as the rodenstock HR designs.

but there will be hard times for the HY6/ AFI, for phase/mamiya and for hasselblad,- the magic thinking in the mf 3d voodoo will not last forever, because more and more "thinking" photographers will demonstrate how good work you can make with systems as the new canons     ( and probably soon the nikons ) can deliver.
« Last Edit: December 03, 2007, 02:33:12 PM by rainer_v » Logged

rainer viertlböck
architecture photographer
munich / germany

www.tangential.de
RobertJ
Sr. Member
****
Offline Offline

Posts: 597


« Reply #39 on: December 03, 2007, 04:08:36 PM »
ReplyReply

The focus points are all different.  All of these tests are invalid.  Bleh.
Logged
Pages: « 1 [2] 3 4 ... 6 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad