My article comparing high-end 35mm digital to medium format, titled The Ultimate Shootout, generated considerable discussion. I'd have been surprised if it hadn't. Based on some comments I've read online I've managed to gore several people's sacred cows.
Most of it has been on web discussion boards while some has been in e-mails. Much of it though has been from readers who have come to similar conclusions.
I have created this page so that some of the comments that I have received as e-mails, both pro and con, can be read by others. Only correspondence in which the writer has something worthwhile to contribute — whether pro or con — will be added. I'll leave out the "you're stupid and don't know what you're doing", or the "Ya, right on!" type messages. There are enough of those on the discussion boards.
I will be brief, but let me first thank you for your web site and its contribution to photography... I enjoy the site and the Video Journal enormously.
Several things on your site have converged to prompt this email.
First, your image of the Snake River overlook at the Tetons. This location also works well at sunrise. One of my favorite images was taken there, with a 4x5, on a very cold SUMMER morning... (looks much better than this, but I did a quick compress to GIF!)
What is interesting about this image is that it was the first 4x5 image I had drum scanned and digitally printed in a LightJet. What amazed me was that a 20x24 inch print at 200dpi was dramatically sharper than any enlarger print I had ever made. I began to realize at that point that 4Kx5K pixels were enough for the eye and that what mattered more than pixel count was the numbers IN the pixels... that grain and noise and edge contrast, not sheer resolution were far more important in the impression of sharpness.
As a seismologist, I work with both wave theory and digitization theory so I began to have a good look at the frequency content, MTF, of all aspect of the photographic process. It became immediately clear that in most cases, certainly in 35mm and probably all the way up to 4x5, the film was limiting contrast more than lenses, and edge contrast, not some Rayleigh resolution criterion, is what the human eye/brain system perceives as sharpness. The more I looked at actual digital images, the more I became convinced that the number of pixels needed for stunning images wasn't anywhere close to the numbers that one would assume if clinging to the traditional measures of resolution in lp/mm. It also became apparent that the reason "resing up" worked so well is that the human eye/brain doesn't miss fine detail that was never captured in those extra pixels, it simply appreciates the tonal smoothness and contrasty edges that are preserved if the process is done correctly. Add in the noise of film grain and the impact of the digital revolution sunk home. I decided at that point (about 3 years ago) that at about 6 MP, digital would equal 35mm, at about 12 MP, digital would equal MF, and (woe is me) by about 24 MP, 4x5 would be equalled. At that point I sold my MF equipment and clung only to my 4x5.
Your tests over the last 18 months have tracked my suppositions through 35mm with the D60 and now to 6x7 with the 1Ds. I suppose it won't be too long and you will demonstrate that the new 22MP Sinar sensor equals 4x5. I can't afford the Eos 1Ds, or the Kodak ProBack, and certainly not the Sinar back, so I am hoping that I can still buy 4x5 transparency film for another 5 years or so until I can afford to buy something in the 10-20MP range.
Which brings me to my final thought. Once I realized that 20MP digital sensors weren't going to be physically very large, about the size of 6x4.5 (the Sinar is a tad smaller than I predicted) I began to wonder if Hasselblad hadn't made an enormous mistake in discontinuing the ArcBody. It seemed to me that the ArcBody with some Scheider or Rodenstock digital lenses and a 20MP digital sensor was the view camera to end all view cameras. Now that Sinar has the sensor, Hasselblad doesn't have the ArcBody anymore... oh well, I couldn't afford that either...
Hope you weather the storm of protests that are bound to follow from folks who look at numbers more than pictures.
Glenn C. Kroeger
Associate Professor of Geosciences
San Antonio, Texas
I just read your 1Ds vs. medium format article. Compelling, to say the least. This backs up what I just noticed with my own eyes, as I went shooting with a 1Ds and a Hassy system last weekend with a friend. Armed with only a Polaroid 120 scanner, the results were quite similar, and have caused me to get thinking about my current equipment. I sold my 4x5 gear last week (yes, all of it), and placed an order for a 1Ds. I should have it next month via my local reseller.
Just read your test results and have posted on the thread in DPR.
You might remember a few months back that I wrote
to you with regards my own findings and the D60 vs MF saying "where has
all the resolution gone then"(referring to MF) I also wrote to you and
went on to buy the 220 vacuum back for the Contax trying to get mf quality to
show it`s head above digital but alas I like you have sold all of my MF gear
and settled on just one camera "1Ds"
One tends to doubt ones self when finding results that shouldn`t really be and although I only used a Nikon 8000, in the end I came to the conclusion that the 8000 was all that I could afford and certainly could not justify £100 scans. So ! the results from an expensive( for me) scanner had to compare to the digital ability of the D60 and 1Ds and at A3 size ,printed and mounted on the wall digital has won the day, every time.
One more quick point, I see no point in members of the forums champing at the bit and wanting to see the results from $300 scans( although I see why you have to do just that) because 99.9% can not afford them and would never consider them as part of their every day work flow. No! we have to compare the results we are getting with the equipment that we have or can afford against digital otherwise it would be like saying " my VW is not as fast as that Formula One car" if you get my point.
The good thing that has come out of your findings is that the man in the street can get Formula One performance from a VW within certain size limits and I for one thank you for for having the balls (excuse the French) to put yourself on the line once again to help discerning photographers get the results they want at the right price.
With your permission I would like to put a link to your report on the MF digest where I took a severe beating recently for daring to suggest that digital is easier and faster with comparable results as opposed to film. Some die hards will never take their heads from the sand.
Thanks for reading and like you I`m looking forward to developments that are coming.
I just read you article comparing the 1Ds to scanned MF film.
First off, let me say this about that -- I made the same decision about four weeks ago and sold my complete MF outfit to justify the purchase of the 1Ds. I hesitated on the 1Ds due to price alone -- but when I realized the 1Ds body effectively replaced my entire Contax 645 MF system, AND that I could use all of my existing Canon SLR lenses, I was hooked. I have Imacon scans of my 35mm and MF negatives, and while they are indeed superb, frankly they do not give me anything above an original 1Ds image. Add upscaling via PS in 10% increments, and personally I think I get BETER results with the 1Ds than I do with the Imacon scans. In short, I justify the extra expense in film-cost savings and system-cost savings in MF alone.
Furthermore, I compared output to many of my prized LF (4x5) images. I found that upon close inspection -- the kind of close inspection I was subjecting the digital image to -- that the LF images were lacking in absolute resolution. They have tonal range in spades, but frankly come up soft in a micro-inspection of detail.
In summary, the 1Ds effectively replaces both MF and LF systems in my photography -- It rocks! Now I have a 35mm replacement in my 1D, and a MF replacement that uses the same lenses in my 1Ds.
Life is good!
Good morning. I agree completely with your article. I've shot 35mm, 4 x5, and panoramic 617 for years. I've had the D30, then the D60, Now it's the D1 and the EOS1Ds. I got the 1Ds at Thanksgiving and have shot all my assignments with it. If I do a 4 x 5 shot , or a panormainc layout, I did the same series with the EOS1Ds to keep some comparitive notes. I got interested in the digital side about 4 years ago, since I also shot Canon 35mm, with most of the L series lens. I also do my own large format printing on the espon 1280, 2220 and the new 7600. The testing methodology and printing series, along with the drum scans are very similar to what you'd done. I can also do a very nice 24 x36 with a some USM and it is very difficult to see the differences between what was being done with a drum scan vs. a worked raw file. I just recently sold my 4 x 5 system, and also sold off my EOS1V bodies. Haven't shot film in about 7 months, and the digital side - whether anyone want to believe it or not - is truely hard to beat !
It can only continue to get better. Anyone who doesnt believe it, should rent a high end digital for an upcoming assignment, and then take a long, serious look at the results. If they dont come to a similar conculsion - they will be passed by their competitors, becasue of their lack of acceptentacnce of where the market is really going.
Your recent comparison of the 1Ds and medium format are fully supported by my own testing. While I did not take the step of sending my test shots off for drum scans, I did show my samples to those who had done so with their medium format chromes. They were astonished not only at the overall quality of the 1Ds images but also at the detail available in very large prints.
Here is an 8.74 x 5.85 inch crop from a 37.5 x 56.44 inch 300 dpi image taken with the 1Ds at these settings: RAW; ISO 100; 1/125; f8.0; and sharpness level 0. The lens was a 24-70mm f2.8L. As the sole light source, I used a 550EX inside a Photoflex medium softbox. The only changes made to the original file were a small amount of USM and a one-step upsizing using the Lanczos 8x8 interpolation method available in Picture Window. No fancy stair-stepping actions were used here, so it could have been much better. But I wanted to see the quick and dirty results. If you look closely at the crop, you can just make out the reflection of the softbox in her glasses. Of course, had I taken the time to balance the main light with some fill, the overexposed highlights could have been eliminated altogether.
The full-frame image is here: http://members.cox.net/willt/1Ds2470Sample/eyedetail_fullframe.jpg
The crop is here: http://members.cox.net/willt/1Ds2470Sample/eyedetail_crop.jpg
By the way, I got similar results from my 70-200mm f2.8L IS and 16-35mm f2.8L. As for the 24mm 3.5L TS-E, it’s good but not in the same league as the others. I hope Canon decides to upgrade this lens.
This and other tests convinced me that film was no longer in my future--the output from the 1Ds is way better than good enough.