1Ds2 versus 5D

[jani]

But he also knows that ACR cannot extract the maximum detail from a 1Ds2 file.  Love the color, love the workflow, but the detail simply aint there.  However, it seems it CAN extract all the detail present on a 5D file -- or at least Pixmantec's Rawshooter does not glean much more from a 5D file than ACR does.  But on a 1Ds2 file, there is no contest, the 1Ds2 files come alive.  Download the freebie Rawshooter if you don't believe me.

Now that's interesting, I have missed this fine point earlier.

Do you know how this might be with e.g. the 20D, 1D2, Nikon D2X and D200, or if anyone else has tested that in a reproducible manner?

I'm so incredibly uncomfortable with RSE's user interface that I cannot perform a comparison between them for the first two mentioned models myself.

[dwdallam]

I agree that if you want to compare the quality of the 5D to the 1D MKII that you should take the same picture with the same lens on both backs.  Use the same software to process say a 20x30 image. Have the two images printed on the same printer. Now what you can say is this:

Given X software with X lens and X processing rules using X printer, the 5D Compared to the DII is X when viewed with my eyes.

That result may or may not mean that the DII isn't much better because what I am hearing here is that ACR can't even process a II file properly!

So when do you think Adobe will correct this--along with offering a 64 bit multithreaded PSCS?

[ErikKaffehr]

Hi!

There is one more factor and that is the antialisaing filter in front of the sensor. The D1S II has a couple of advantages in this sense:

1) They can probably afford a higher price AA filter (they are expensive)
2) As they are closer to resolution limits of lenses they can have a weaker AA-filter

The AA filter acts a "low pass filter" techically speeking. Unfortunately it not really possible to make an optical low pass filter

Erik

It is a big difference, but it's a big difference on a huge print that would be far larger than any giclee type printer could handle.

On my 19" monitor with screen resolution of 1280x1024, a 400% enlargement of the entire image equates to a print size of approximately 17.5 feet wide. If I view such a print from a distance equal to the print's diagonal, then such differences that I see viewing the 400% crop from a distance equal to my monitor's diagonal, disappear completely. In fact, I only have to step back about 12ft from my monitor for those differences to become undiscernible.

Now I know people have a habit of walking up close to a print, whatever the size. I some times do this with billboards to check out the size of the ink dots (just out of curiosity   ), but I submit, if these two 17.5ft wide prints were side by side on a gallery wall and you wanted to compare micro detail, by the time you'd walked over to the other print, you'd probably have forgotten the precise nature of the micro detail you'd been examining just 30 seconds ago. (Maybe not   ).

I haven't made any prints of these crops yet, as Tim suggests, but at 67% enlargement on my monitor, the full size image is as big as my Epson 7600 can handle without borders on all 4 sides (ie, 36x24").

At this degree of enlargement (67%) the differences clearly visible at 400% are no longer discernible. It's my experience that what I can't see on the monitor, I can't see on the print (excluding subtle hues of color that might be outside the gamut of the monitor but inside the gamut of the ink and paper).

By the way, that 14% increase in resolution that the 1Ds2 is supposed to have is a theoretical calculation that can only be realised with a perfect lens. It represents an increase in sensor resolution as opposed to system resolution. In practice, the resolution increase will always be smaller and herein lies the difficulty as I see it.

The 30% increase in pixel count, of the 1Ds2, appears fairly substantial and creates the impression that the resolution increase should be substantial. However, the resolution increase in terms of lp/mm along one dimension is a theoretical maximum of 14% and probably no more than half that in practice.

In my view, some people incorrectly interpret this small increase (say 7% in practice) as an indication that current lenses are simply not good enough to justify any further increase in pixel density. My interpretation is, a 30% increase in pixel count is simply too trivial to make a worthwhile difference. You need to double pixel count to get an improvement that hits you in the eye. Such improvement, however, would be overkill for the printing of uncropped images at normal or average print sizes, but great for those wanting to get the most from their longest lenses, or for those occasions when an interesting composition can be found by heavily cropping an image.
[a href=\"index.php?act=findpost&pid=69613\"][{POST_SNAPBACK}][/a]

[Jack Flesher]

Now that's interesting, I have missed this fine point earlier.

Do you know how this might be with e.g. the 20D, 1D2, Nikon D2X and D200, or if anyone else has tested that in a reproducible manner?

I'm so incredibly uncomfortable with RSE's user interface that I cannot perform a comparison between them for the first two mentioned models myself.
[a href=\"index.php?act=findpost&pid=69778\"][{POST_SNAPBACK}][/a]

I have never converted 20D files on anything except RSP so I cannot comment.  

On my 1D2, RSP definitely gave a more detailed final file than ACR, enough so that I saw no appreciable gain in OVERALL resolution from the 5D -- a fact I cannot explain since the two cameras have the same pixel pitch.

I don't have enough experience with Nikon's digital cameras or NEF's to comment on them.

Once you learn RSP/E interface, it is a batch dream, however very different than ACR's workflow.

[Jack Flesher]

So when do you think Adobe will correct this--

[a href=\"index.php?act=findpost&pid=69789\"][{POST_SNAPBACK}][/a]

I think it is pretty clear they just did that with their acquisition of Pixmantec, the developers of RSE/P.  I am willing to bet we will have that raw conversion technology in CS3's ACR and/or the final version of Lightroom.

[Ray]

But he also knows that ACR cannot extract the maximum detail from a 1Ds2 file.  Love the color, love the workflow, but the detail simply aint there.  However, it seems it CAN extract all the detail present on a 5D file -- or at least Pixmantec's Rawshooter does not glean much more from a 5D file than ACR does.  But on a 1Ds2 file, there is no contest, the 1Ds2 files come alive.  Download the freebie Rawshooter if you don't believe me. 

[a href=\"index.php?act=findpost&pid=69746\"][{POST_SNAPBACK}][/a]

That's not the impression I get when converting the 2 files under discussion using RSP. If anything, I would say the additional detail extracted by RSP is greater in the 5D image than it is in the 1Ds image.

In the images below, I've used the default zero setting of the detail extractor and no sharpening for the RSP conversions. For the ACR conversions I've used the default sharpening at 25, which I think is fair because RSP's detail extractor appears as though it's a type of sharpening.

[attachment=785:attachment]

[Jack Flesher]

That's not the impression I get when converting the 2 files under discussion using RSP. If anything, I would say the additional detail extracted by RSP is greater in the 5D image than it is in the 1Ds image.

In the images below, I've used the default zero setting of the detail extractor and no sharpening for the RSP conversions. For the ACR conversions I've used the default sharpening at 25, which I think is fair because RSP's detail extractor appears as though it's a type of sharpening.

[attachment=785:attachment]
[a href=\"index.php?act=findpost&pid=69825\"][{POST_SNAPBACK}][/a]

And here we go...

Detail extraction is not really a type of sharpening -- it is in fact more a type of noise reduction, or more accurately a lack of it.

To get the most from a 1Ds2 file you have to add sharpening during the conversion step -- so if you turned sharpening off in RSP for the 1Ds2 file you were handicapping it and and there is even more detail to be gleaned.

Sharpening off in RSP is not the same as setting the slider to zero -- it is the toggle button on the batch conversion page.  For the 1Ds2 file try leaving sharpening on and set the slider to -10.  Detail extraction at zero is fine and will render a smooth image, though I would use +5 for a well lit ISO 100 file and get even a bit more detail out of the image even though it might have a slighter harsher look.  

So I guess one could claim it is fair to leave sharpening off in RSP, but only if you want to handicap the RSP conversion to make it look closer to the ACR conversion.  Again, I don't think you will see a significant difference between the two converters on the 5D file when both are optimised -- but with the 1Ds2 files you most certainly will.

[Ray]

Jack,
There are a lot of options. I didn't try RSP sharpening at -10 but at zero I could see halos. In any case, it's a very subjective impression to claim that the improvement in one image is greater or less than in another image. Having looked at other parts of both images, it's clear that the 1Ds2 really does produces more detail than the 5D even at 100% on screen (which still represents a print about 65" wide on my monitor.

I'll try some more conversions with -10 sharpening, but the images below give a definite edge to the 1Ds2. I assume that focussing would have been on the same spot with both cameras and accurate   .

[attachment=786:attachment]

[Jack Flesher]

it's clear that the 1Ds2 really does produces more detail than the 5D even at 100% on screen (which still represents a print about 65" wide on my monitor.

[attachment=786:attachment]
[a href=\"index.php?act=findpost&pid=69834\"][{POST_SNAPBACK}][/a]

Yes it is quite clear -- and as you know, it always has been to me  

But keep in mind even though your onscreen resolution is effectively a 65" print, you will still see a significant difference in a 16" or 20" print.  If you don't believe me, take the images you processed and cropped to form the above set and size them all to 16" at 300 PPI, crop those same portions out from each and print it on a gloss or luster photo paper with a good profile.   If you do, I suspect you'll clearly see the advantages to the 1Ds2 RSP over the 5D RSP or 5D ACR in the resultant prints and will also clearly see the advantages in the 1Ds2 RSP file over the 1Ds2 ACR file (assuming you sharpen in RSP).  Though you probably won't see much of a difference between the 1Ds2 ACR, 5D ACR and 5D RSP files.

[Ray]

Yes it is quite clear -- and as you know, it always has been to me  

[a href=\"index.php?act=findpost&pid=69838\"][{POST_SNAPBACK}][/a]

Not so fast, Jack. It looks as though Bill Janes was right. The tests are not valid, but not for the reasons he gave. Focussing appears to be out.

Below is another comparison from the same 2 images showing the 5D provides better detail. What can I say! Cough! Cough!  

Perhaps the differences between the 5D and 1Ds2 are truly so small that the best attempts of a group of experienced professionals cannot provide sufficiently accurate focussing to reveal such differences.

Hope I'm not embarrassing anyone here   .

[attachment=789:attachment]

[Ray]

I suppose it's now incumbent upon me to look at the other shots at f11, f16 and f22. I chose the f8 shots to compare because some folks seem very concerned that Canon lenses are not good enough. All good 35mm lenses are sharper at f8 than at f11 to f22. The dollar bill was my first point of comparison because I understood from the article that this was the point of focus.

I've just done a series of conversion in RSP using Jack's recommended settings of -10 for sharpening and +5 for detail extraction. You might find the comparisons below revealing. Draw your own conclusions.

[attachment=790:attachment]                         [attachment=791:attachment]


[attachment=792:attachment]                         [attachment=793:attachment]


[attachment=794:attachment]

[Eric Myrvaagnes]

Not so fast, Jack. It looks as though Bill Janes was right. The tests are not valid, but not for the reasons he gave. Focussing appears to be out.

Below is another comparison from the same 2 images showing the 5D provides better detail. What can I say! Cough! Cough!   

Perhaps the differences between the 5D and 1Ds2 are truly so small that the best attempts of a group of experienced professionals cannot provide sufficiently accurate focussing to reveal such differences.

Hope I'm not embarrassing anyone here   .

[attachment=789:attachment]
[a href=\"index.php?act=findpost&pid=69861\"][{POST_SNAPBACK}][/a]

Well, Ray, nobody can claim that you aren't persistent!
Throughout this thread I've been lurking, and getting my hots for a 5D. Even with this new bombshell, it looks like a pretty good upgrade from my 10D. Thank you Ray, Jack, Bill, and any other contributors to this lively thread.

Eric

[Jack Flesher]

Great effort Ray.  Problem is that you start suffering from diffraction effects on the 1Ds2 after f8 and seriously so at anything over f10 on it.  The 5D's larger pixel will be somewhat more forgiving in showing this and the effects of diffraction will begin to equalise the two cameras regardless of processing parameters.

And unfortunately we run into the problem of many lenses not being up to the sensor at apertures wider than f4 or f5.6, so to glean the maximum potential from the 1Ds2 one really needs to shoot it primarily between about f4.5 and f10 -- a fairly narrow band, but fortunately (for me) one that integrates well with the types of imaging I mostly do.

IOW, if you shoot a lot under f4 or over f11, in most cases you won't see much detail advantage in the 1Ds2 over the 5D regardless of lens used.  

Cheers,

[Ray]

Great effort Ray.  Problem is that you start suffering from diffraction effects on the 1Ds2 after f8 [a href=\"index.php?act=findpost&pid=69881\"][{POST_SNAPBACK}][/a]

I know, I know, Jack. But let's get down to practicalities. Perfect focussing is an ideal rarely achieved. In a studio set-up, I guess you can achieve perfect focussing on the model's eye lashes. Everything else is mediocre or totally out of focus.

Show me such some shots, comparing the 1Ds2 with the 5D (or other equivalent shots). Focussing is critical.

[Jack Flesher]

Focussing is critical.
[a href=\"index.php?act=findpost&pid=69886\"][{POST_SNAPBACK}][/a]

Never said it wasn't

I am also on record saying if you don't use a tripod, the 1Ds2 will give little over the 5D.  Or if you don't use the best lenses, the best exposure...  

Keep in mind the test images you are playing around with were 1) shot using the camera's built-in AF and 2) they were shot in a studio environment and the main subject is only about 1-1/2 meters away.

Re #1: How accurately the camera in question AF's is important.

Re #2: Focus is significantly more critical at closer distances than it is at longer distances.  

,

[Ray]

I agree completely. So show me the results. I thought Michael's tests would be definitive, but it seems they aren't.

Show me some definitve results, Jack, so we can put this matter to rest.

[Jack Flesher]

I agree completely. So show me the results. I thought Michael's tests would be definitive, but it seems they aren't.

Show me some definitve results, Jack, so we can put this matter to rest.
[a href=\"index.php?act=findpost&pid=69889\"][{POST_SNAPBACK}][/a]

We already have them -- just use Michael's f8 shots.

,

[dwdallam]

We already have them -- just use Michael's f8 shots.

,
[a href=\"index.php?act=findpost&pid=69892\"][{POST_SNAPBACK}][/a]

This proves that for any tests other than hardware type tests that photozone does, it's a crap shoot. On photozone, they tested they 24-70L against the 28-70L and the lab tests showed that teh 28-70 was technically superior to the 24-70L. In Michael's tests, it's obvious that the 24-70 has an advantage, but where we go again with all the variables in an uncontrolled environment.

I would say take the best picture in a studio that you can given a specific lens with both backs. Upsize them to 20x30 at 320 ppi and then print them using the same printer. Let your eye tell you what's up.

[Ray]

I would say take the best picture in a studio that you can given a specific lens with both backs. Upsize them to 20x30 at 320 ppi and then print them using the same printer. Let your eye tell you what's up.
[a href=\"index.php?act=findpost&pid=69898\"][{POST_SNAPBACK}][/a]

Having examined various sections of the f8 shots on the screen at magnifications ranging from 100% to 400%, I've got a pretty good idea of what to expect on a 24x36" print. Some parts of the print will appear sharper on the 1Ds2 print. Some parts will appear sharper on the 5D print, and some parts will appear equally sharp on both prints.

As we move up (or down depending on your biases) from f8 to f22, some of the differences will merge till at f22 they are virtually indistinguishable, but not completely indistinguishable as in the following examples which have been sharpened just a tad with Focus Magic. As I see it, we've got almost the same difference here, even at f22, at what must be close to the actual point of focus.

[attachment=795:attachment]

[jani]

This proves that for any tests other than hardware type tests that photozone does, it's a crap shoot. On photozone, they tested they 24-70L against the 28-70L and the lab tests showed that teh 28-70 was technically superior to the 24-70L. In Michael's tests, it's obvious that the 24-70 has an advantage, but where we go again with all the variables in an uncontrolled environment.

Also add the other tests that show a slight superiority ni the favour of the 24-70, and you'll have general mayhem.

But this is what clinches it for the 24-70:

 - moisture/dust seals
 - availability (you can buy it new)
 - bokeh (according to some people, anyway)

Oops, that was a digression.

On the general point, I agree with you. I really miss Photodo, but even Photodo didn't take everything into consideration (bokeh, for instance, is completely missing, and vignetting isn't insignificant either).