Michael: a question

[lisa_r]

QUOTE (pegelli @ Mar 12 2009, 11:19 AM)
I also liked the piece in there called "beyond resolution". I always thought I was crazy seeing more details from FF or MF sensors vs. APS-C even in web images downsized to 800 or 900 pixels on the long end (0.5 MP  )

It seems there is more to sharpness and detail impression than resolution alone, local contrast and actuance play a role as well and apparently more of those are saved in a downsized picture than one might think.

I ran into a friend at PMA in Las Vegas, and over a coffee he commented that he found that even on the web he thought that he could see a difference between my P65+ shots from Antarctica and those from the 25 Megapixel Sony A900.

This is the case with medium format digital images today. Even though a 21-25MP DSLR can produce exceptional prints, one can clearly see the difference that medium format makes, sometimes even at the web's low resolution.

Even some very savvy technical types in the industry don't have a firm explanation for what it is we're seeing. The best explanation that I've been able to come up with is what I like to call micro-contrast. What we appear to be seeing in large format film and medium format digital (especially from 39MP and up) is the ability for the system to differentiate tiny differences in luminance values, tonality and colour.

Micheal, I posted this query elsewhere but I am not sure you saw it, as I didn't see a response. Just curious as to how the above remarks about seeing differences between these cameras on web jpegs fit in with your report that respected pros could not reliably tell the difference between 13x19" prints from the Canon G10 vs. the p45+ back prints...

Thanks.

[BernardLanguillier]

Micheal, I posted this query elsewhere but I am not sure you saw it, as I didn't see a response. Just curious as to how the above remarks about seeing differences between these cameras on web jpegs fit in with your report that respected pros could not reliably tell the difference between 13x19" prints from the Canon G10 vs. the p45+ back prints...

For what it is worth, my view is that it mostly boils down to DR/tones. Michael's G10/P45 test produced the results it producted because the scene had so much DR in it that even the P45+ could not handle it. On less extreme scenes, the additional DR of the back/DSLR woud IMHO show a lot. I don't believe that the following image could have been produced with a G10 for example.



As far as resolution goes, I do also feel that even small images show some qualities telling about their original resolution. When I downloaded the full resolution image of the P45+, it didn't appear close to what I am normally getting with high end DSLRs. I don't believe that that image was shot optimally but even then, it is a scene that will not show big differences because it is mostly filled with emptiness between a few more of less sharp trees. Shoot a subject with many details all over, and the difference between a P45+ and a G10 will be very visible even in small prints IMHO.



So all in all, Michael's provocative G10-P45+ article was mostly aimed at showing that the gap can be fairly small in some circumpstances, but I don't believe he ever claimed that a G10 will be impossible to distinguish from a P45+/high end DSLR accross the board. Please correct me if I am wrong Michael.

Cheers,
Bernard

[Ray]

Hey! We all make mistakes   . I think Michael must be now regretting that comparison.

From my perspective, the comparison was flawed because the DoF was not equalised. The P45+ print was identified eventually because of its shallower DoF. If Michael had used an appropriate F stop with the P45+ to equalise DoF, which would have been something like F25, the G10 image would possibly have been sharper, and those photographic experts viewing the prints might have identified the G10 shot as the P45+. How does that grab you?  

[BobDavid]

Comparing a G10 to a P45+ was a boneheaded move. I could make a Kodak 126 Instamatic photo compare favorably to a photo shot with a Phase back if the subject matter fit.

[Ray]

Comparing a G10 to a P45+ was a boneheaded move. I could make a Kodak 126 Instamatic photo compare favorably to a photo shot with a Phase back if the subject matter fit.

Well, that's a bit extreme. The subject matter was not ideal to demonstrate DR or a silky smooth tonality, but it was a fairly typical landscape and the point that, for an A3+ size print, the benefits of a P45+ would have been unnecessary in these circumstances, is worth making.

[michael]

No, I have no regrets at all. It was a provocative and fun little experiment. Unfortunately some people misinterpreted it. (And for those that think I have some sort of "sweetheart" relationship with Phase One, imagine how they responded the next morning to my favorable comparison of one of their backs to a $500 P&S).

To my mind the comparison simply showed that there's a whole lot more to image quality than some people think. It has a huge amount to do with people's expectations and experience as well as shooting conditions, presentation and so forth.

I just finished an open house at my gallery last night where about 300 people came through between the opening on Thursday evening and the weekend. The 27 prints hanging were typically sized at about 22X28", matted and framed to 28X34". With a couple of exceptions the images were a mix of Phase One P65+ and Sony A900 from Antarctica, Arizona and Utah.

People generally fell into one of two groups - those who instantly could identify which prints had been shot with which camera, and those that couldn't and who would frequently get them wrong when trying to guess which was which. I should note that many of those who couldn't tell the difference reliably were experienced pros and some from the publishing and advertising industry, just as with the P45+ and G10 experiment.

All this tells me is that different people have differing abilities to discern image characteristic differences. The other day one of my all-day students astonished me by pointing out gamut differences between prints on different papers that I simply couldn't see, and I think of myself as having a refined eye for these things. He could see things that I just couldn't.

So, let's not jump to too many conclusions. Keep an open mind and an open heart.

Michael

[Snook]

Michael you really have a HIGH tolerance level and always seem reserved with your comments..
I applaud you for that .. Really.

Snook

[lisa_r]

Interesting Micheal, thanks for your thoughtful reply.
One day I will visit your gallery.

[ivan muller]

Michael you really have a HIGH tolerance level and always seem reserved with your comments..
I applaud you for that .. Really.

Snook

why not snook? It costs nothing, shows one maturity and its really not that hard to be humble....


regards, Ivan

[Snook]

why not snook? It costs nothing, shows one maturity and its really not that hard to be humble....


regards, Ivan

Not really sure what you mean by your post.
I was complementing Michael. I think being humble is harder than you think..
It was a genuine comment..

:+}
Snook

[smhoer]

[quote name='ivan muller' date='Mar 16 2009, 09:39 AM' post='268195']
why not snook? It costs nothing, shows one maturity and its really not that hard to be humble....


     

[feppe]

This question comes up every so often, so I decided to run a quiz a while back. Conclusion: the respondents were no better at picking MFDB shots based on web-size JPEGs than by guessing. The quiz itself is not functional anymore.

[Harold Clark]

No, I have no regrets at all. It was a provocative and fun little experiment. Unfortunately some people misinterpreted it. (And for those that think I have some sort of "sweetheart" relationship with Phase One, imagine how they responded the next morning to my favorable comparison of one of their backs to a $500 P&S).

To my mind the comparison simply showed that there's a whole lot more to image quality than some people think. It has a huge amount to do with people's expectations and experience as well as shooting conditions, presentation and so forth.

I just finished an open house at my gallery last night where about 300 people came through between the opening on Thursday evening and the weekend. The 27 prints hanging were typically sized at about 22X28", matted and framed to 28X34". With a couple of exceptions the images were a mix of Phase One P65+ and Sony A900 from Antarctica, Arizona and Utah.

People generally fell into one of two groups - those who instantly could identify which prints had been shot with which camera, and those that couldn't and who would frequently get them wrong when trying to guess which was which. I should note that many of those who couldn't tell the difference reliably were experienced pros and some from the publishing and advertising industry, just as with the P45+ and G10 experiment.

All this tells me is that different people have differing abilities to discern image characteristic differences. The other day one of my all-day students astonished me by pointing out gamut differences between prints on different papers that I simply couldn't see, and I think of myself as having a refined eye for these things. He could see things that I just couldn't.

So, let's not jump to too many conclusions. Keep an open mind and an open heart.

Michael

This is very well put. There are many factors involved in perception: preconceived expectations, experience and education in the particular field, and probably even more important the innate individual differences mentioned here which are hard to quantify. To draw a musical analogy, some people can detect the most minute imperfection in a performance, while others are tone deaf and can't detect one note from another, or differentiate between a major or minor key.

When digital photography was new, I perceived a higher level of quality in prints than was justified by the technical facts. I eventually realized that I making judgments based on previous experience. There is no grain in digital prints, and many years of experience had me equating grainless prints with larger film sizes, hence higher quality.