Hi Anders,
We had a lengthy discussion on this forum about the Tim Parkin test and I was the person starting that thread.
I am most impressed by Tim's test. That said it was not really an MFDB vs. DSLR test. DSLRs were included in the test but played a marginal role. The test was essentially a resolution test and didn't look into DR.
I had no issue with the test itself. But at the resolution tested the small sensor devices fall apart.
What I have seen in Tim's test mainly was that large format film could outperform even the IQ180.
On the other hand, Tim has his own high end drum scanner and he scanned at 8000 PPI. I guess that it takes some time. Tim has bought his scanner at very reasonable price.
I guess that no one really questions that a 80 MP MFDB has better image detail than a 24 MP DSLR. The questions may be:
- What do I gain for spending 20 kUSD on a low end MF digital (lenses included) over a high end DSLR?
- What do I gain buying an old back (like 16-25 MP backs) over a high end DSLR?
- If the MFDB is superior to a DSLR at which print size will the difference be significant?
Questions like above are not the focus of Tim's article.
Regarding fairness I don't agree that Tim's test is the only fair test published. I'd say that all tests are fair as long as:
- The test is not designed to put a certain device at disadvantage
- The test is properly made
- All parameters are disclosed
For instance, the analogue vs. digital tests I made are fair in my view. I have done everything I could to set up a level playing field. I even had some very high resolution drum scans made with the friendly help of Dominique Ventzke, at 6000 PPI. The only omission is that I have tested the equipment I have. Would I have a higher performing lens the results may have been better, but I have not. You know, I have invested many hours of work in those tests, just choosing an image for involves scanning 4-5 focus bracketed images at highest resolution, a couple of hours of work.
One issue worth mentioning is that we have seen in Tim's that even 4x5" could match the IQ180, at least in some sense. Some earlier tests indicated that P45 performed similar to Velvia scanned at 2000 PPI. This discrepancy came probably from the fact that the authors of those tests used 2000 PPI on their 4x5 scans. So the tests was not really intended to find out how much performance could be eaked from 4x5" but to see if the backs were good enough to replace 4x5" in the authors workflow. This applies to the "Great MFDB shootout of 2006" by Reichmann, Atkinson, Cramer and Sanderson:
http://www.luminous-landscape.com/essays/back-testing.shtmlThe enclosed image is from my test and shows a small crop at 6000PPI. In this image the Sony Alpha is actually at disadvantage because of camera vibration (due to enabled Antishake).
Best regards
Erik
For anyone interested my test is here:
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/59-sony-alpha-900-vs-67-analogue-round-2?start=1I really should have posted following yesterday and perhaps many of posts above would not have been.
http://www.onlandscape.co.uk/issues/lgb-0028/
That test was posted here on LuLa by Tim Parkin recent. It is per my knowledge the only fair test that so far has been published and that involves "Medium Format Digital Vs. Large Sensor DSLR" but also film.
There are obvious clear differences that per my impression puts an end to arguments.
Very interesting point. If I was a professional photographer I may well be using a DSLR as my main tool, simply to get the job done quick, done, money, higher ISO, and since suffice image quality for the publishing requirements. However for times when I saw need to push for a competitive edge or simply saw desire or need to provide better image quality, I would grab MFDB. Money wise, indeed MFDB is expensive, but so are constant upgrades of DSLRs.
Best regards,
Anders