Digital vs. MF Scanned Film Tests???

Discussion in 'Medium Format' started by r_scott_steube, May 20, 2007.

  1. I love to shoot Ilford PAN F Plus (6x6 to 6x9), then typically scan using Nikon
    CS 9000ED for printing 24x28+ inch (or larger) wall art. Perhaps one in a
    thousand image capture warrants ICG drum scan by a photo service.

    What I like to find is a very thorough online article that tests/contrasts MF
    Negative Scans vs 35mm D-SLR vs MF Digital Backs for Black and White Wall Art.
    I not have much luck finding this online; could someone point me in the right
    direction? Many Thanks!
  2. If you can't find it with Google, it's probably not there. The prevailing wisdom is outresolves high-end 35 mm DSLR beats MF film. That's the conventional wisdom, which I don't agree and don't believe, not when you look at the resolving power of film and top-end Hassy/Zeiss MF lenses.

    The difference between 35 mm and MF is that tripod is used a lot more by MF shooters, hence the chance of getting full mileage out of the film is much better when camera shake is not in the equation.
  3. Maybe this link can help some. For what it is worth I get exelent 16"x20' from my 645. You should have no problem geting exelent enlargements from you cameras specialy the 6"x9" I would say as good as (mega pixels = more money cameras/backs) I have found out that for amateurs used medium formant is the way to go right now if you like to large enlargments an do not have the money for state of the art digital equipment.
  4. If you are near a large city, I suggest you rent a 1Ds MKII or digital back for your camera.
    Rentals are available for these things...especially in major markets. Go out and shoot the
    same subject with your 6x6 and 6x9 and then do the same with the digital. Print both as
    you normally would and compare. Then YOU will be the one making the judgement. This is
    critical, because different people see differently and have different tastes. <P>In my own
    experience, digital gives a cleaner looking print when the sensor is approximately equal in
    size to the area of the film. I use a Leica M8 and a Leica DMR, both are 10mp with no AA
    filter and use the best glass around. They compare very well with Canon's 12 and 17mp
    cameras. Up to 16x20, they compare pretty well with medium format too. Beyond that, I
    still think medium format film gives a better look. I just recently printed 30x30 on a 9800
    from imacon scans of a Hasselblad neg and the results are astonishingly good. <P> But
    with black and white, as you ask, I think film looks better. I have no idea what it is, but I
    have not been able to reproduce the look of a well shot black and white negative with a
    color conversion in photoshop. I much prefer the look of Acros, Tri-X or Neopan 400 to
    that which I can get with a digital conversion. I am wholly ready to admit that it might be
    my skills, but regardless, film looks better to me for black and white.
  5. From the little experience I have, I think you will not find what you want from a DSLR, no matter how many pixels, because - and I am really guesing here - the tonal transitions are limited by the 8 or 12 bit recording, and the dynamic range is often too compressed for the subject. To my eye, a (good) DSLR is good enough in colour to make me abandon colour 135mm film. In fact I shoot colour portraits with a Fuji S3 with Zeiss ZF lenses, and the results are remarkable, but in B&W I stll prefer film.

    Against MF in colour it is a tough call, but the digital DSLR is better for some things and worse for others.

    I think you should explore the MF digi backs - the best have a huge DR, and are already 14 bit. Apparently the best quality imaginable comes from the scanning backs if you can put up with all the hassle.
  6. Here's a sample of Fuji Astia (chrome) from a Hasselblad, scanned on an Imacon:
    1:1 crop</ a>
  7. I don't think you will find a direct comparison. But with your parameters, I believe small format DSLRs will lose clearly. MF backs may seriously challenge your film, but I really can't comment on that due to me lacking experience with MF backs.
  8. "I love to shoot Ilford PAN F Plus (6x6 to 6x9)" ... then just keep loving it. To retain your images in both analogue and digital gives you the best of both worlds, and safe.

    Rod's example of good scanning from 6x6 film is impressive enough for me.

    My only interest in digital is the convenience of putting artwork on-line with minimum fuss. My Hasselblad images, certainly the B&W work is a pixel-free zone, and I'm not missing out on anything.

  9. I have seen a few examples of tests comparing MF with dslr (but not MF digital back and not in B&W wall art). One of them was just a few examples shots...on this forum. I cannot find the link right now, but if you search you may come up with it. Another was done on

    I have also seen criticisms of the above comparison. I think, as Stuart says, the best thing would be to conduct your own side-by-side comparison, keeping as many variables the same as possible (lenses of roughly the same quality, same "equivalent" focal lengths, etc.), though others may have some better tips in that regard.

    Below is my personal experience with MR and digital in landscapes. I thought I'd include it in case there was info that might be helpful to you:

    I like to shoot landscapes from a certain viewpoint overlooking a river gorge with forest flanking it. With digital SLR's (including the Nikon D50 that I own--with Tamron 17-55, and rented Rebel xti and Canon 30d-- with Canon 10-22mm), the trees in the far did not show up well...they consistently looked mushy, with artifacts, as you zoomed in to 100%.

    However, when I rented a Mamiya 645, the same area was noticably sharper (though I shot at dusk and the whole thing had a tremendous blue cast with the Velvia 100f so was unusable). To be fair, I did have some shots with the Mamiya that looked softer than I would have liked-- of a waterfall with background vegetation-- and the waterfall is blown out a little. It all looks like mush except for the waterfall (could be due to user error). That brings me to another point: Digital has superior dynamic range when compared to Velvia film, which is the most widely used for landscapes.

    In another example, I shot a waterfall with a mossy background on two occasions, one with the Mamiya 645afd 55-100 zoom and the Canon 30d with. Both with high DOF. The detail in the Mamiya shot was clearly superior.

    I think this difference in detail advantage may be virtually nill when compared to full frame digital. However, there is also the issue of tonality and tonal gradation. I have read anecdotal reports--on this forum--that they are better with medium format.
  10. I still haven't sold the hasselblad collection because I feel the photos are still better compared to the Canon line or the top of the line Nikon D2X. Sharpness as well as color. Color can be corrected to a great degree using tools like Photoshop, but not the sharpness. I have a very good scanner, 4800X4800 for MF but it still doesn't cut it when you print directly from a neg or a positive and make an enlargement from a medium format.

    I'm hoping to buy a digital back for the Hassy's this year.
  11. Jay - digital may well have better (if better = longer/wider) dynamic range than Velvia - but so does pretty much every other color film currently made. :) In fact, I'd hazard to say that a good number of color negative films have more dynamic range than digital sensors, and in B&W, I'd say it's not even close.
  12. i've used a nikon D70 and scanned 6x7 negs from my RB67 and graflex century (with nikon nikkor) and the scans blow it away, by alot.
  13. I've had similar results as Jay mentioned above. There's no way my D80 with any lens can equal the detail I get with my old Mamiya Press 23 and 100mm lens. I use it with a 6x9 back and the detail I see in things like foreground grasses is amazing. For other subjects the 10 MP of my D80 is often good enough.
  14. more recent luminous landscape test: 645 vs P45 vs 1DsmkII
    Looking at the pics on here 645 velvia drum scanned beat the 1DsmkII in sharpness but not grain/noise. Shame there was no 6x7 test included.
    This test also contradicts this test, though I think the results were evaluated on 13x19" prints.
    P67 vs 1Ds
  15. </a></a>
  16. There are not a lot of comparisons of the sort you're looking for; I don't recall seeing any tests that specifically compared B&W performance. The most thorough comparison (in terms of number of cameras tested) that I recall is the Luminous Landscape "Measuring Megabytes" test (Canon 5D, 1Ds, 1DsII, 3 Phase 1 backs from 22-39 MP, and a 6k Betterlight scanning back, along with 645 and 4x5 film):

    You can take a look at these images and see what you think. When I look, I see what my own testing has also shown: a 12 MP DSLR is in between 35 and 645 in terms of resolution and perhaps equal to 645 in perceived image quality when both are shot at ISO 100-200. (At ISO 1600, the DSLR is hugely superior to 645, but that's irrelevant for landscapes.)

    My experience with a Canon 5D, 645, and 6x7 doing B&W and color landscapes leads me to grab the 6x7 when I want to make a print larger than 16x20. It's to some extent a judgment call, since the 5D produces essentially noise-free images at ISO 100-200; I still find that the higher resolution of 6x7 is much more important than the difference between the slight grain of ISO 100 film and the undetectable digital noise of the 5D at the same ISO.
  17. This all depends on what you shoot. I find even 10mp is enough for portraiture and weddings. When I'm shooting landscape work, even at 16x24 the 5D lacks something in microfine detail....especially in B&W. For B&W work, I still prefer 6x7 Delta 100 and Pan F to my 1Ds Mk2. Large format slaughters the DSLRs without effort!
  18. good Less, finally somebody put and end to this BS about 35mm vs medium format. If somebody dissagrees, send a post in the canon's forum. :) if you can :)
  19. read "send a post to the"
  20. This is a comparison that came about simply because me and a friend happened to be taking pics and we had different cameras...... These are 100% crops that from photos that were taken from about the same kind of spot, and then resized to 511 pixels, on a Canon 5D with a 20mm lens (non-L) and a Hasselblad SWC/M with Kodak E100G, scanned on an Imacon 949. Both have been sharpened a little. The SWC scan ended up offering a final print size a lot larger than the 5D file, which I guess is the real thing friend thinks he can push the 5D to about an A2 print, and I think I can push the SWC to about A0. R
  21. We should also bear in mind the different format ratios between MF and Canon's full-frame sensor. 645 and 6x7 have 4:3 and 5:4 aspect ratios, whereas full-frame has a 3:2 aspect ratio.

    Now if we chop the full-frame sensor down to the more commonly used 4:3 ratio, we lose one-eighth of the frame area. This makes Canon's 16 megapixel sensor only 14 megapixels, and their 12.8 megapixel sensor only 11.4 Mp in direct comparison to 645. Compared with 6x7 the numbers are worse, with 16Mp being chopped down to only 13.3Mp when utilising a similar aspect ratio.

    Don't get me wrong. I think that Canon's offerings are steps in the right direction, but to be serious contenders against MF and LF, the pixel count has to rise dramatically, and it would help if they left behind old Oscar Barnack's haphazard choice of aspect ratio. Give us a 24x32mm sized sensor with around 30 megapixels, in a 35mm SLR sized body, taking existing lenses, and everyone will be happy.

    Come on you camera companies! We know the technology exists. Just stop drip feeding us your old low-tech, low-res techno-tat in a cynical effort to part us from our money on a never-ending treadmill of upgrades.
  22. Early in the 20th century, in the search for a more portable camera format, 35mm
    movie film was adapted for use in still cameras. At first, there was no standard
    frame size, but eventually, the industry settled on 36 x 24mm. Great rangefinder
    systems were developed by Leitz (Leica) and Zeiss (Contax), Canon and Nikon,
    among others. In the latter half of the century, professional 35mm photography
    was dominated by SLR systems developed by Canon and Nikon. An immense array
    of lenses was a powerful component of these systems. When the first DSLR
    cameras reached professionals and consumers in the late 1990s and early 2000s,
    their image sensors were not as large as the standard 35mm frame, causing some
    consternation. Photographers started pressing camera manufacturers to offer 36 x
    24mm image sensors.
    Medium format 120/220 film is 6cm wide. The smallest common image sizes
    on this film are 60 x 60mm (2 1/4 inches) and 60 x 45mm (2 1/4 x 1 5/8 inches).
    Excellent, if relatively limited, lens systems have been developed for these formats.
    While some vendors are calling their very expensive 49 x 36.7mm image sensors
    ?full-frame,? they are not, at least in terms of medium format. This isn?t to say that
    there are no true full-frame medium format sensors available. For the record, and
    as a matter of interest to tech-heads, Fairchild Imaging?s CCD 486 has a 61.44 x
    61.44mm image area and 16 megapixels. It is included in liquid-cooled format with
    the Peregrine 486 (extreme sensitivity under low light conditions) and Condor 486
    (X-ray and electron imaging) scientific cameras. It is offered separately in three
    grades and in frontside and backside illuminated configurations. The cheapest
    (frontside, grade 3) is $16,000; the priciest is $95,000 for a backside-illuminated
    grade 1.
    Fairchild?s CCD 595, for ?advanced scientific, space and aerial reconnaissance
    applications,? has a rather spectacular 81 megapixels on its 80.64 x 80.64mm image
    area. The sensor alone, sans camera or associated electronics, costs approximately
    $100,000, depending upon packaging and application. The point of mentioning this
    exotica here is to frame the discussion of, and to establish the cost of, advanced
    sensor technology and to give a clearer sense of Canon?s position in the
    marketplace. This paper will demonstrate the excellence and the value of Canon?s
    full-frame technology as seen in the context of both less expensive and more
    expensive units. (source: canon)
  23. The difference between Canon 5D and a scanned medium format slide or negative is very slim. 6x9 probably still outperforms 5D full frame sensor. The most important question is post-processing: what are you planning to do with the photos: Project them? Print them? If so, how? Traditional darkroom process? Inkjet? I myself just recently sold my D5 and went back to Hasselblad, I simply enjoy the process and prefer the "Look" of traditional b/w.
  24. FWIW, here is my experience as someone who shoots digital (D2Xs), medium format (Bronica SQ-A), and 4x5, prints both digitally from drum scans and in the darkroom, and has had the opportunity to critically review the output from every DSLR out there during my workshops. There is no question that a good drum scan from a quality 6x6 or 6x9cm film original shot on high resolution film with quality optics outresolves all DSLRs. This difference will show up clearly in the large prints you like to make. My understanding is that even if a 22MP full-frame DSLR were released, MF film would still outresolve it simply due to the larger imaging area. In real world shooting conditions, the resolving power of most 35mm lenses on a 24x36mm sensor is maxed out in 16-18MP range anyway, so the lenses would have to be redesigned and manufactured to higher tolerances to see any real image quality improvement. Given how ridiculously inexpensive used MF film gear is these days, and that 120 b&w film costs about $3/roll, I personally find film a very compelling solution for b&w work (what a surprise!). My Bronica SQ-A system was purchased recently with six PS lenses, two backs, and a non- metered prism for $1200. If you compare this to the cost of a modern digital MF system that will actually compete with drum scans from MF film, you'll find that the change you save will buy a vast amount of film, processing, and drum scanning! Develop your own film and you will save even more and have more control over the final tonal characteristics of the film. By the way, this points to the fact that resolution isn't the only issue. B&w film has a vast ten-stop (+/- depending on development) dynamic range compared to digital sensors, and that full range can be pulled out of the film by a drum scanner, but would be lost with digital capture. The other advantage of film in b&w work is that you can still print it in the darkroom. Hand-made darkroom prints are taken far more seriously by the serious fine art market than digital prints of any kind. Also, I was just out photographing the other day with the Bronica in a wind-storm along the shores of Mono Lake with alkaline salt water spray flying around. Most people would think twice before doing that with a $20K+ digital back. In my case, if a lens had gotten saturated with water and trashed, I would have been out maybe $200. If I drop a film back in the lake, I can pick up a replacement in great condition for $100. Then there is the power issue. The attached image was made from a seven minute exposure on film, and was one of four brackets I shot between four and fifteen minutes as the light faded. A friend who was using a DSLR at the time burned through his battery in no time trying to do long exposures. I still prefer the look, quality, and practicality of film, and the hybrid approach offers the best of both worlds.
  25. i agree 100% with justin, $33.000 digital back outdoor it's not i bright idea :)

Share This Page