645D - Are my expectations unreasonable? Near-edges very soft at infinity [imgs]

[Rob C]

Detail.

To resort to the anecdotal, derided in these pages some time earlier, I look at my own experience. Nikon and Hasselblad gave much the same detail in the circumstances in which I found myself working. The general max. size I ever needed for most things printed perfectly well onto 10"x12" paper, from either system (as seems to be true today with digital prints on A3+ from either type of digital camera). For larger work, such as full- or half-page press ads and fashion show and shop displays on 40"x60" paper, either camera was as useful because the working distance - the viewing distance - was relatively great and detail was resolved much the same way. What was rather different, though, was the tonality.

Tonality.

I assume the dxo question about it was posted in jest, but if not, it displays one of the current problems besetting photography today: snappers abrogating experience for the illusion of mechanically/electronically measured values which you can't always get; I wonder about the dxo rating of Caravaggio or even dear old Vinnie van G - can't be measured? can't be valid, then. Interesting take on life, my friend.

The definition of tonality that I grew up with was fairly simple: it's the quality of the visible and uninterrupted continuum of a grey or a colour that is not broken down by the failures of the medium upon which it is created. Tonality and grain are the usual suspects here, and both are the unavoidable factors that produce the quality of tonality as well as its failure. Judgement of tonality depends on a trained, experienced eye that understands what's possible and where it is either maintained or lost. That said, even the untrained eye should be able instantly to detect the differences in tonality in two similar sized prints from two very differently sized negatives or positives. The trick, of course, is having the skill to make judgement calls when confronted with a single print and no comparator.

Rob C

[Rob C]

The level of detail needed for a given task and the level of detail available are quite different animals.

If billboards are your thing then viewing distance is applicable, rather less so for anything purporting to be fine art anything.

Quite so, and that's where user discretion has to rule. To spend or not to spend, is one system 'better' for me than another? Which brings the circle back to the point and question of whether using mf at all is realistic for some shooters, not just economically but from the working perspective too. It's much the same as the old one about running a Ferrari in Malta.

Rob C

[ErikKaffehr]

Hi,

The problem it seems is that there are many samples of the 55/2.8 which don't impress testers, specially in corners and edges, so it may not be a sample problem with your lens but a more specific problem with the 55/2.8. It seems that there are many other Pentax lenses that work fine, but possibly lack AF and definitively ultrasonic drive.


Best regards
Erik

Hmmm. Well anyhow, to return to the original topic from the OP -

* No, your expectations are not at all unreasonable.

* There is probably absolutely nothing wrong with your camera (or the one before). It is the 55mm lens which seems to be the problem.

So request that your dealer swap the 55mm for a better lens.

John

[hjulenissen]

Tonality.

I assume the dxo question about it was posted in jest, but if not, it displays one of the current problems besetting photography today: snappers abrogating experience for the illusion of mechanically/electronically measured values which you can't always get; I wonder about the dxo rating of Caravaggio or even dear old Vinnie van G - can't be measured? can't be valid, then. Interesting take on life, my friend.

The definition of tonality that I grew up with was fairly simple: it's the quality of the visible and uninterrupted continuum of a grey or a colour that is not broken down by the failures of the medium upon which it is created. Tonality and grain are the usual suspects here, and both are the unavoidable factors that produce the quality of tonality as well as its failure. Judgement of tonality depends on a trained, experienced eye that understands what's possible and where it is either maintained or lost. That said, even the untrained eye should be able instantly to detect the differences in tonality in two similar sized prints from two very differently sized negatives or positives. The trick, of course, is having the skill to make judgement calls when confronted with a single print and no comparator.

Rob C

I mentioned DXO and fair side-by-sides because they are both "fair" within their respective limits. Measurements are not inherently "bad", but one should know what they measure, and have skill to interpret (extrapolate) them. That is not always the case, and I suspect that is why some peopleare so negative towards measurements. I find comparision to measurements of painters to be falsely constructed for this discussion, and dont see
any point in discussing them further.

Statements about trained experienced eye, skill to make judgement call etc may well be correct, but they also open to all kinds of "emperrors clothes" issues.

If there are real, visible, repeatable differences in what you call tonality, surely someone must by now have made fair comparisions between a 5Dmk2 and a 645D, tweaking raw developer settings for visual similarity (documenting what they did), and showing us clearly the benefit of MF?

-h

[Rob C]

I mentioned DXO and fair side-by-sides because they are both "fair" within their respective limits. Measurements are not inherently "bad", but one should know what they measure, and have skill to interpret (extrapolate) them. That is not always the case, and I suspect that is why some peopleare so negative towards measurements. I find comparision to measurements of painters to be falsely constructed for this discussion, and dont see
any point in discussing them further.Statements about trained experienced eye, skill to make judgement call etc may well be correct, but they also open to all kinds of "emperrors clothes" issues.

If there are real, visible, repeatable differences in what you call tonality, surely someone must by now have made fair comparisions between a 5Dmk2 and a 645D, tweaking raw developer settings for visual similarity (documenting what they did), and showing us clearly the benefit of MF?

-h

Exactly my points:

a. why would you accept something you can't dispute if it spoils your argument;

b. measurements are sooo scientific, but art and its appreciation never can be, and photographic tonality no less so. I can only conclude that if you, personally, can't distinguish tonality as either good or poor, then you have little photographic experience which indeed renders this correspondence futile.

But hell, it's a new year, let's not all get bogged down in more verbal trash! I, for one, shall try not to add to what's already choking the word.

Rob C

[hjulenissen]

a. why would you accept something you can't dispute if it spoils your argument;

I am not sure what you are referring to, but you brought painters into the discussion. I think that the quality of painted art cannot be measured, so do you, so there seems to be nothing to discuss.

If you claim is that if the quality of painted art cannot be measured, then the quality of photographic equipment cannot be measured, then I disagree as those are two very different things.

b. measurements are sooo scientific, but art and its appreciation never can be, and photographic tonality no less so.

I am sure that there are really good photographers out there that dont read dxo measurements, but dont know what medium format is either. The point is: if you state that medium format makes better art than other formats, then you are allready stepping out of the art-making and into technical explanations.

Making bold statements about how the world works (something like "MF has much better tonality than anything else"), then hiding behind the "I am making art, so anyone questioning my conclusions are electronic nerds"-argument does little to advance my or your understanding, or the state of the art.

I can only conclude that if you, personally, can't distinguish tonality as either good or poor, then you have little photographic experience which indeed renders this correspondence futile.

I make no claims about my photographic experience, but the sentence above tells me more about you than it does about tonality. If it cannot be seen, cannot be smelled, cannot be heard, cannot be explained, then it probably only exists in the head of the true believers. I have no intention of becoming a religious person, I joined this forum to improve my photography skills and understanding of my equipment, and this is not helping.

-h

[Joe Behar]

Making bold statements about how the world works (something like "MF has much better tonality than anything else"), then hiding behind the "I am making art, so anyone questioning my conclusions are electronic nerds"-argument does little to advance my or your understanding, or the state of the art.I make no claims about my photographic experience, but the sentence above tells me more about you than it does about tonality. If it cannot be seen, cannot be smelled, cannot be heard, cannot be explained, then it probably only exists in the head of the true believers. I have no intention of becoming a religious person, I joined this forum to improve my photography skills and understanding of my equipment, and this is not helping.

-h

Just because YOU can't see, smell, hear or explain something does not mean it does not exist. I'm sure that I can't tell the difference between an "OK" and a great violin, but Yitzhak Perlman can. I SHOULD be able to measure the thickness of the wood, the vibration qualities and such and completely understand a violin and tell Mr. Perlman which one will be the best instrument for him, but I can't.

Things like what Rob said will not help you understand the technical aspects of you equipment, but if you think for a moment it wont improve your photography skills, I'm sorry to tell you that you are incorrect.

[tsjanik]

Smoody:

I have a 645D, but not the new 55mm.  I am in the process of testing the 17 645/67 lenses I own and have not seen anything remotely as bad as you've posted, in fact, the older lenses are doing quite well.  So unless you've drawn two bad bodies, it appears to be a lens issue  I've posted some results representative of what I'm seeing here where I am comparing the 645D to 645N film scans:
http://tsjanik.blogspot.com/2011/01/blog-post_9035.html
The first image is a crop from the 645D, no post processing (120mm macro); the second is the full image taken on Astia; the third is a 645N crop with PP (detail has been lost in the jpeg conversion). 
The old 55mm is quite good and inexpensive, but manual focus.

Tom   

[hjulenissen]

Just because YOU can't see, smell, hear or explain something does not mean it does not exist. I'm sure that I can't tell the difference between an "OK" and a great violin, but Yitzhak Perlman can. I SHOULD be able to measure the thickness of the wood, the vibration qualities and such and completely understand a violin and tell Mr. Perlman which one will be the best instrument for him, but I can't.

My point is that if mr Perlman cannot distinguish between an OK and a great violin blindfolded, then he is probably fooling himself subconsciously if he claims that one is "better" than the other. If he cannot produce two recordings blindfolded of the two where he himself or others can pick out which is which, then choice of violin does not seem to matter wrg acoustical properties (as long as he is going to do sighted concerts it may still matter because having an expensive one may make him feel better and thus do a better performance, but that is probably not the explanation that was given in the first place).

I started out with a simple question: is it possible to measure tonality using DXO, or to show us the effect in a fair side-by-side set of images. Neither of you have answered this in a sensible manner, instead you have attacked measurements in general, and focused on my photography skills (or lack thereof). I am sad to say that that kind of reaction does not make me trust your conclusions or statements.

Things like what Rob said will not help you understand the technical aspects of you equipment, but if you think for a moment it wont improve your photography skills, I'm sorry to tell you that you are incorrect

I am sorry, but this is nonsense. Arguments of the tonality of MF cameras wont make me take better pictures. If you think that, then we are indeed living on very different planets. Going out taking pictures may make me a better photograph, studying the work of others may make me a better photograph, getting critizism for my work may make me a better photograph. Thinking about composition, what is aesthetically pleasing, what draws my attention to a scene in the first place may make me a better photograph. But nit-picking about one of the supposed benefits of an expensive camera technical feature (tonality vs Sensor-size) that no-one are willing to show examples of, is almost certainly not making me a better photograph.

-h

[John R Smith]

I think that a great deal of this endless wrangling over MF versus 35mm completely misses some very important points.

If you take a photograph of a well-lit, perfectly framed subject in ideal conditions, the difference between an MF camera and something like a 5D2 or D3X in the final print will probably be utterly minimal. You would have to enlarge the shot to 36x24 or something to see any MF advantage. However, where the MF has a huge advantage and will save your arse is when things go wrong, as (in my case at least) they do so often. If the subject is a difficult one, with either too much contrast, or too little, and needs lots of post-shoot manipulation, you will find that you can push and pull the MF file around in LR or ACR or whatever far more than the 35mm one. Those big MF files will take a lot of messing with, and still look good. The huge other advantage of MF is when your framing was, shall we say, less than optimal. I can easily crop down to just 25% of the original image area and it still looks just fine printed to A3. In fact, sometimes I can get two good pictures (and different) out of one MF frame.

So for someone like me, who so often totally fouls up the shot, MF is absolutely the bee's knees   And I'm not too proud to admit it.

John

[tho_mas]

Tonality have always been a vague term to me. Is it something that one can read out of dxo measurements, or something that can be clearly pointed to in a side-by-side comparision of images from cameras that "good" tonality vs "bad" tonality?

+1

If there are real, visible, repeatable differences in what you call tonality, surely someone must by now have made fair comparisions between a 5Dmk2 and a 645D, tweaking raw developer settings for visual similarity (documenting what they did), and showing us clearly the benefit of MF?

Here are RAW-file comparisions (I think Canon 1DsIII / P65+ & D700 / P65+):
http://www.phaseone.com/de-DE/Downloads/Sample-images.aspx

You can play around with the files.
You need Capture One Pro to do so… there's a free 30day trial on the website.
Note that some images already contain settings (i.e. not "preset" settings; they are adjusted for certain purposes. There was an article these files refer to, hence the adjustments applied to the captures. But I don't know whether or not the article is still available somewhere... )
Also note that in terms of tonality a "comparision" of course is seriously limited by the quality, gamut and linearity of your monitor.

[ondebanks]

Why on earth would you choose to shoot portraits with a 55mm? Surely its natural role is for landscape and architecture, just the subjects which will suffer from poor sharpness across the frame.

John

But bear in mind the crop factor: 55mm on the 645D gives the same diagonal FOV as an 80mm lens on a 6x6cm film camera...and a hell of a lot of portraits were shot down the years with 80mm standard lenses on all those Rolleis, Hasselblads, Bronicas etc. (Even though the perspective afforded by a 120-180mm lens on 6x6cm would be preferable).

That's why Pentax chose the 55mm focal length for their new 645D standard - it's the whole "a 'standard/normal' lens must give a 1 radian FOV" thing that has come down to us through photographic history. Personally I prefer either a long-standard or a moderate wideangle; if I had a 645D these would be the 75mm and 45mm lenses respectively.

Ray

[ondebanks]

As for 6x7 format lenses beating 35mm format lenses, that's into science fiction. The whole problem associated with people making the step up from 35mm cameras to medium format (top quality in both cases) was that they assumed that they were multiplying 35mm quality by the film size factor, when the truth is that 35mm formats are such superlative peformers precisely because the designers only have to cover a small area with the 35mm format, and simply can't achieve the same degree of quality over the larger area mf presents... The only real advantage from mf has always been that of tonality. Period.

Rob C

Rob, that just ain't so. 6x7 format lenses will hardly 'beat' 35mm format lenses, but they can certainly equal them. There is/was real equality between the on-film resolution of the best 35mm and MF lenses. 80 lp/mm is quite readily attainable from both formats - more 35mm lenses might deliver it than MF lenses, but plenty of MF lenses still do. Look, for example, at Chris Perez' tests. Some 35mm and MF lenses go to 120 lp/mm and beyond (including some of the MF lenses in that very test I linked to - yes, even 6x7 ones), but there are precious few films which can deliver that resolution at any sort of usable MTF contrast. All that lovely extra aerial resolution is lost in film emulsion scattering and grain effects.

This being the case, one really is "multiplying 35mm quality by the film size factor", or near enough. That's why MF film images generally appear so much more detailed than 35mm film images (as well as having better tonality in a given size of print, which I agree is indisputable). Honestly, I've never seen this dismissed before.

In fact, it is the same reason why MFD is a much harder sell than MF film was. One could move from a 35mm Leica to a Mamiya 7 and know that you would be getting 4 times the net image quality (in modern parlance, double the lp/ph in both axes). This is not the case if you move from a FF-DSLR or Leica M9 to a MFD system, which is typically only 2 times the sensor area (only 1.7 times in the case of the 645D).

Ray

[Joe Behar]

I started out with a simple question: is it possible to measure tonality using DXO, or to show us the effect in a fair side-by-side set of images. Neither of you have answered this in a sensible manner, instead you have attacked measurements in general, and focused on my photography skills (or lack thereof). I am sad to say that that kind of reaction does not make me trust your conclusions or statements.I am sorry, but this is nonsense. Arguments of the tonality of MF cameras wont make me take better pictures. If you think that, then we are indeed living on very different planets. Going out taking pictures may make me a better photograph, studying the work of others may make me a better photograph, getting critizism for my work may make me a better photograph. Thinking about composition, what is aesthetically pleasing, what draws my attention to a scene in the first place may make me a better photograph. But nit-picking about one of the supposed benefits of an expensive camera technical feature (tonality vs Sensor-size) that no-one are willing to show examples of, is almost certainly not making me a better photograph.

-h

h,


You answered your own question on the first part, so I won't bother expanding on it any further.

The second part is what I maybe did not make clear enough, and you have misunderstood.

Its not about the "benefits" of one format over another that will make you a better photographer...its knowing the DIFFERENCES.

Some of those differences can be measured, some cannot. Not because they are magic, but rather because they cannot be expressed as simple numbers.  Its almost like trying to measure and define a good image and coming up with a mathematical equation to show why its good.

All I claim, is that there is a DIFFERENCE and untill you see for yourself what the difference is, you cannot decide for sure which system is best for you. Isn't the whole premise the image rather than charts and graphs?

Have a look. You might be convinced, you might not, but I can assure you that either way, you will have learned something, and learning ANYTHING, can make you a better photographer. Please don't dismiss anything as rubbish till you've gone through it and seen for yourself.

[hjulenissen]

I think that a great deal of this endless wrangling over MF versus 35mm completely misses some very important points.

If you take a photograph of a well-lit, perfectly framed subject in ideal conditions, the difference between an MF camera and something like a 5D2 or D3X in the final print will probably be utterly minimal. You would have to enlarge the shot to 36x24 or something to see any MF advantage. However, where the MF has a huge advantage and will save your arse is when things go wrong, as (in my case at least) they do so often. If the subject is a difficult one, with either too much contrast, or too little, and needs lots of post-shoot manipulation, you will find that you can push and pull the MF file around in LR or ACR or whatever far more than the 35mm one. Those big MF files will take a lot of messing with, and still look good. The huge other advantage of MF is when your framing was, shall we say, less than optimal. I can easily crop down to just 25% of the original image area and it still looks just fine printed to A3. In fact, sometimes I can get two good pictures (and different) out of one MF frame.

So for someone like me, who so often totally fouls up the shot, MF is absolutely the bee's knees   And I'm not too proud to admit it.

John

Thank you for your opinion.

I have no objections against resolution, and accept that a good 40 megapixel camera can (given the right lenses, enough light, recording technique etc) resolve more detail than a good 15 megapixel camera. What conditions are necessary for that extra information to be relevant and important is besides the point, which is that the information potentially is there.

My original question was about the more vague "tonality". If I interpret you correctly, you find that MF files can stand more dynamic processing before artifacts like banding, noise, black/white-clipping becomes unbearable, or simply having more information hidden away in places that may not be visible using untweaked files? If that is the case, it should be relatively simple to set up a fair side-by-side test where the MF image is tweaked to "subjectively sane , good looks", and the 5Dmk2 alternative is tweaked to represent the scene similarly, but the 5Dmk2 showing artifacts, lack of shadow detail, or similar flaws. A good measurement of SNR and DR at the given camera settings should (along with e.g. examples of the noise quality) give good insight into the numbers at play?

-h

[hjulenissen]

Here are RAW-file comparisions (I think Canon 1DsIII / P65+ & D700 / P65+):
http://www.phaseone.com/de-DE/Downloads/Sample-images.aspx

You can play around with the files.
You need Capture One Pro to do so… there's a free 30day trial on the website.
Note that some images already contain settings (i.e. not "preset" settings; they are adjusted for certain purposes. There was an article these files refer to, hence the adjustments applied to the captures. But I don't know whether or not the article is still available somewhere... )
Also note that in terms of tonality a "comparision" of course is seriously limited by the quality, gamut and linearity of your monitor.

Thank you. I will try to have look at those files. Do you think that they represent a "fair comparision"? (I note that the URL seems to be that of a MF camera manufacturer)

Of course, the image display chain can be a potential limitation. I am using a new, calibrated Dell 2711 27". I guess that others here will have better displays than myself, one reason why measurements are nice to have alongside real-world images.

-h

[tho_mas]

Do you think that they represent a "fair comparision"?

"fair" is always hard to define when you have to mix formats. Do you shoot the same horizontal dimensions? Or the same vertical? Or the same diagonal? And so on.
But I didn't look at the images in detail anyway... so I don't know.

Here's a comparision of a number of different cameras (645 + 35) that has been posted numerous times.
http://www.flickr.com/photos/dos-chin/sets/72157614936120567/

[ErikKaffehr]

Hi,

Here are my impressions from that test: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/41-phase-one-images-for-download

In short I'd suggest it's a good comparison, except for outdoor portrait. I'd also would prefer a comparison with Nikon D3x instead of Canon 1DsIII and Nikon D700.

The reason I don't think the outdoor portrait is not a correct test is that a normal lens (50/1. is used at full aperture, it's just not sharp. The idea is to have short DoF, but that is better achieved with a longer lens.

It's very clear that DR in P65+ is much better than 1DsIII, it has very clear shadow detail. I have done my owns raw processing using Capture One demo to convert sample files to DNG.

Best regards
Erik

Thank you. I will try to have look at those files. Do you think that they represent a "fair comparision"? (I note that the URL seems to be that of a MF camera manufacturer)

Of course, the image display chain can be a potential limitation. I am using a new, calibrated Dell 2711 27". I guess that others here will have better displays than myself, one reason why measurements are nice to have alongside real-world images.

-h

[ErikKaffehr]

Hi,

I have an issue with tonality, because in my view tonality has everything to do with postprocessing and little with sensor. The sensor itself is a linear device, in essence, and just delivers a signal. Interpretation is done in postprocessing. So I cannot see how MF cameras can have better tonality. If they have better DR, that could be helpful in achieving better tonality.

I have downloaded the files and played around with them. In my view the P65+ was vastly superior to the Canon 1DsIII in DR. Both lenses had an incredible amount of lateral chroma, by the way, but that could be cleaned up in postprocessing.

My findings are here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/41-phase-one-images-for-download

I have done a similar comparison on the Leica S2, using images from "Diglloyd":
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/38-observations-on-leica-s2-raw-images
The emphasis in that article was on aliasing/Moiré. If found it pretty clear that the S2 was impressive in sharpness, but I actually found that the Nikon D3X was better in DR.

Finally I also run a comparison on the Pentax 645D. In this case I made actual prints and looked mainly at resolution. The Pentax was sharper than the D3X. The images I analysed were not appropriate for checking DR, but I guess that the Pentax is pretty impressive in the DR area.
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/51-a-closer-look-at-pentax-645d-image-quality
Some unexpected Moiré was visible. I have seen some aliasing in most MF samples I looked at carefully, so lack of OLP-filtering comes with a price.

To sum it up: the samples I have seen deliver on promise regarding sharpness. I don't think that Leica S2 has better DR than Nikon D3X. Aliasing may be a problem at some times. It seems that stopping down to f/11 on 6 micron pitch sensors can eliminate most aliasing artifacts due to diffraction.

Best regards
Erik

Here are RAW-file comparisions (I think Canon 1DsIII / P65+ & D700 / P65+):
http://www.phaseone.com/de-DE/Downloads/Sample-images.aspx

You can play around with the files.
You need Capture One Pro to do so… there's a free 30day trial on the website.
Note that some images already contain settings (i.e. not "preset" settings; they are adjusted for certain purposes. There was an article these files refer to, hence the adjustments applied to the captures. But I don't know whether or not the article is still available somewhere... )
Also note that in terms of tonality a "comparision" of course is seriously limited by the quality, gamut and linearity of your monitor.

[tho_mas]

It seems that stopping down to f/11 on 6 micron pitch sensors can eliminate most aliasing artifacts due to diffraction.

depends on the lenses. The 65+ bridge shot from that comparision is at f11 and shows moiré (at infinity). I've also seen other f11 shots from a P40+ / P65+ showing moiré. Both are 6µm sensors.
My P45 (6.8µm) shows moiré even at f16... depending on the lens.