Photo printout comparison between MF vs DSLR

<p>Hi, do anyone here have seriously compared the photo printout say 20x24 or display sizes (colour photo)taken by MF gear in films (negative or transparent), direct silverprint or scanned by high-end scanners, with modern High-end, High-resolution's DSLR or even prosumer grade Aps DSLR. What would be the difference? I am very much appreciate for your comments. Thanks. </p>

<p>Aside from the character of film being different than digital, MF cameras still produce images in the 100mb range or more (depending on the negative size). A couple of colleagues and I did an experiment by taking the exact same scene with a Nikon D3 and Mamiya 645. We scanned the 6x4.5 MF slide with a Hasselblad scanner, and our colleague worked in his version from the Nikon. Then they were printed through the same lab to the same 16x20", 20x24", and 30x40" sizes. Even with the 16x20" print, we all agreed the MF result was clearly better, and result was more obvious in the larger two prints. The prints were color, not B&W, and the MF film was Fuji Provia 100. This result would surely be even more substantial with a 6x6, 6x7 or 6x9 MF transparency, obviously.</p>

<p>This is a very difficult topic as the images are different. Here are my conclusions from a number of tests.<br>

1 Black and white done from film in a wet process is still the best. I have scanned it but I personally do not like the way black and white film scans so I generally process it in a darkroom<br>

2 I find that 645 MF (Mamiya Pro / ProTL) slide film when scanned produces results that are as sharp as my 5DII and as photographs I tend to slightly prefer the Mamiya results. I scan on a Nikon 9000 and the Mamiya files are 200-250 MB per shot. I can post some examples but the results are very close between the two methods<br>

3 Moving up to my GX680 (6x8cm) the Fuji scanned results are definately superior to the Canon. However, I suspect a lot of the difference is due to lens performance. The GX680 lenses (with due respect to Zeiss and Leica) are about as good as lenses get - perhaps this is why Fuji makes the H series lenses, similar performance for less money than Zeiss. The image files of the Nikon 9000 are between 300MB (8 bit colour) and up to 650 MB (16 bit colour). <br>

4 Colour negative does not scan as well as a positive so the 645 images are not as good as the 5DII and the 6x8 images are probably slightly better at 100 ISO but I suspect will be worse at 400ISO. I have really only compared 400 ISO at 35mm as I do not use many print films in MF bodies - especially over 100 ISO.<br>

Let me know if you are interested in seing some sections of images - I have not managed to sucessfully post full images as when I compress files down from these sizes strange things happen. To get a 650 MB TIFF file down to a 2 MB JPEG is not what i would think of as a sucessful process.</p>

<p>You should also be aware that the scanner is an important aspect of the process and can be very time consuming. Big files are also require a lot of storage and are sow when you edit them.<br>

All this is to say that for top quality large enlargement landscape photographs my 5DII with L series glass cannot compete with the Fuji GX680 but gets close to the Mamiya 645s. My guess is that the DSLRs will get a little bit better but that lens performance will become their limiting factor. In addition the Fuji has full front movement with all of its lenses (except the 50mm).</p>

<p >Michael and Philip, thanks for your information. </p>

<p > </p>

<p >I do not intend to raise another debate between film and digital. But just wish to know whether a latest High-end DSLR have the ability to compete the image quality produced by MF film nowadays. </p>

<p > </p>

<p >I owned Hasselbald camera with several lens for more than 15 years. Perhaps photography is just my hobby but I desired very much in image quality therefore I upgraded my equipment to MF. However, as the price of high-end DSLR going down, I have the intention to buy one like Philips's to replace my MF, not only benefits of ease of use, save processing time and cost and also the availability of ultra long telephoto lens. </p>

<p > </p>

<p >As long as silver-base enlargement is not available in my local Lab, I really miss the nature film look result. Now, they could only provide service of scan-and-print for both negative films and slides. I wonder whether the final results may not better than a modern high-resolution DSLR with more than 20M effective pixels. </p>

<p > </p>

<p >I wish to find my way to go and looking forward to have your comment and happy to share with your great experience. Thanks.</p>

<p>"What would be the difference?" - Actually, not a lot!<br>

With colour, a full-frame DSLR gives quality very close to what can be obtained from MF film. In fact the DSLR will easily beat slide film when it comes to dynamic range. I agree that there's still nothing to touch film and silver-gelatine prints for B&W, but for colour the DSLR is pretty close to MF. And when you have to wind up the ISO sensitivity past 400 then the DSLR is the clear winner.</p>

<p>I think the main difference is that most MF users always use prime lenses, whereas DSLR snappers tend to have a rather crappy zoom strapped to the front all the time. A full-frame DSLR fitted with a good prime can produce quite amazing image quality. Far superior to anything that 35mm film ever produced.</p>

<p>You also have to consider the overall system cost. Good MF lenses aren't at all cheap, and neither is a decent MF scanner. The cost of a decent scanner alone can outweigh that of a pro-quality DSLR, and if you're going to farm out the scanning and printing to a lab - well, what's the point?</p>

<p>I also find it interesting that the previous posters are all using a hybrid film/digital workflow, and not comparing a DSLR to MF film traditionally enlarged.</p>

<p>Joe,</p>

<p>Could you show us that and how a full frame DSLR clearly wins from MF film?</p>

<p>Philip,</p>

<p>Colour negative scans lots better than positive.<br />Colour negative is the better film in all respects, except suitability for being used in family slide shows. ;-)</p>

<p>Lens performance will not be the limiting factor.<br>

Not as long as sensor technology does not make a giant leap forward, and get rid off, for instance, the need for bayer and other patterns.<br>

Aliassing too will keep hanging round digital technology's neck like a millstone for the foreseeable future.<br>

Then there is noise, an increasing problem with higher pixel densities, i.e. increasing pixel count within a fixed, small sensor format (there still is a marked difference in quality between full-frame DSLRs and MF digital capture).<br>

So no need to start worrying about lenses.</p>

<p>Joe - as I stated I find that MF film will beat a good DSLR - I use all L series glass on a 5DII which is Canon's best sensor. You are correct that Digital has a wider dynamic range than film - especially one like Fuji Velvia but I prefer the colours of Velvia. I scan because that is pretty much the only way to print from slide film these days - Cibachrome etc... all went years ago. I kmow of no lab that prints from slide film without scanning. I almost always use a wet process for black and white and the rolls of colour print that I shoot usually get printed in a darkroom. I do not always print my own colour film as I do not shoot a lot and setting everything up for a colour process is rather time consuming given the temperature sensitivity of the materials.</p>

<p>Q.G. I find that I get better results from scanning slide than print. When I shoot print I tend to use Fuji Pro 160 (or 400 with some 35mm rolls). I also shoot Kodak Portra 160 NC (sometimes VC). While I use a Nikon 9000 Scanner I also use the Nikon software and find that I get better results with slide film. If you have any suggestions let me know. I am surprised that you beleive that lens performance is not a limiting factor for DSLRs like the 5DII. I know that if I put a cheap lens on it (e.g. a 35-70 EOS lens that the kids have) the camera shows all of the lens failings. Similarly I suspect that with low contrast targets and at the edge of the frame the lens perfomance limits the final image quality. I do not find that this is the case with the GX 680. If you look at the 5DII sensor you have 156 pixels per mm which is better than the high contrast resolution of most 35mm lenses. To get the equivalent situation on the 78mm wide Fuji negative the lens would only need to resolve 72 lines per mm. Am I miss understanding you or are you suggesting that lens resolution is not a limiting factor for a dense DSLR sensor.</p>

<p>There's room for a few more issues in this one. </p>

<ul>

<li>Those people volunteering their experience seem to have used the Coolscan 9000 for scanning film. Now I accept that this is a decent scanner- in fact I used to own one- but I do wonder whether this is the best way- and not just a good affordable way- to present the film half of this assessment to its best advantage. Would the conclusions be similar using drum scans for example?</li>

<li>There is a great deal of this comparison that needs to be considered subjectively. Certainly in most people's hands there is very often a difference in look between film and digital. You can't quantify everything. </li>

<li>My own experience tells me that there are certain types of photograph that are much easier to get with a digital slr than with medium format film. Given the choice between a decent image and potentially no image or a poor image on film I'll take the former. Equally the reverse is of course true and certainly when I look at fall foliage photographs with a lot of tiny detail I much prefer those made on film to those made on digital. So what sort of photgraphy you want to do and how will influence the answer. </li>

<li>Its hard to bracket the different ways of printing from film together. On the one hand I still, like others here, prefer the look and feel of a traditionally printed b&w photograph. But with colour, I strongly prefer the scan and print route over a traditional enlargement in terms of the size I can get, the sharpness and detail at larger sizes, and the controlability/repeatability of the process. </li>

<li>Its also hard to suppose that all prosumer and above DSLRs are the same and that all MF film is the same. The answer to your question might be very different comparing a Holga or Seagull's output on a fast film to a top of pro-line Nikon or Canon than you'd get from comparing say a Fuji 6x8 using Velvia with a 30D or similar. </li>

</ul>

<p>So if the OP was hoping for a nice clear-cut answer - 20 plays 18 maybe- then he is going to be sorely disappointed. </p>

<p>David - your assessment is spot on and matches my own. to me one of the big advantages of MF is that it makes you slow down and think more. I know that you can do the same with a DSLR but somehow you tend not too. While a drum scanner produces slightly better results than the Nikon 9000 from the few professional scans I have had done I found the differences small and not worth the extra cost (for me a 6x8 Imacon scan costs $70 CDN after tax). If you are only planning to buy a single camera system than a full frame DSLR is probably the best and most versatile option. I shoot a range of systems from rangefinders via micro 4/3 to SLRs (digital and Film but all full frame) through Mamiya 645 to Fuji 6x8. At some point I will probably add Large Format just for completeness. I was answering the question from my perspective where a big MF camera system will still just beat a high end DSLR (Canon even suggests that the 5DII has their best sensor - although I imagine the Nikon D3x is better and the replacement for the 1DsIII will be better). That said I take a 35mm system out the most often and Digital is now at least half of my outings with a camera. The Fuji and to a lesser extent the Mamiya are special event cameras - the things you take for a dawn shoot for example - they are not general purpose. It should also be remembered that to get the results from MF film is a much more time consuming process than downloading a card. I still love to shoot film (I think part of it is waiting to see the results) but for most practical uses (LF Landscapes aside and B&W aside) you can get very similar results much faster with a top DSLR.</p>

<p>I think it is not particularly useful to criticize a hybrid digital and "analog" print process. The question wasn't about a digital vs. negative/tran to print, it was about the image quality. The problem comparing an all-analog print to one from a digital source is that it introduces too many variables and a comparison between image quality is not really relevent. There is a lot of speculation about the difference between formats, and that's why we decided to determine the difference on our own. It was pretty clear to us.</p>

<p>Philip,</p>

<p>A quick, incomplete reply for now: the 5D II produces an output that would equate to 156 P/mm.<br>

But that doesn't translate to a physical resolution of 78 lp/mm (which actually is a low figure) because of Bayer patterns and dumbing down filters (a.k.a. anti-aliassing filters).<br>

Don't worry about lenses. ;-)</p>

<p>Regarding Cibachrome-it's still being made under the name Ilfochrome and there are people printing it commercially.</p>

<blockquote>

<p>I think it is not particularly useful to criticize a hybrid digital and "analog" print process. The question wasn't about a digital vs. negative/tran to print, it was about the image quality.</p>

</blockquote>

<p>I don't think thats true. I think the OP's question is about print quality and that the quality of the original image is but a part of this. Therefore it is not only reasonable but essntial to take into account the complex interaction of variables that produce prints. It is entirely possible that what starts off as the best original image makes a poor print if sub-optimal decisions are made about what processes are used to print and if necessary scan. There's much more to this question than lp/mm, though of course that is part of it.</p>

<p>Speaking of which, Philip's point on scanners. My own subjective impression was that I'm happy making prints of say 18" sq from a Coolscan but that for a larger size (which the OP specifies) then I'd be a little happier with a scan from an Imacon or a drum scan than from the Nikon. I won't be providing "evidence" since that view is subjective, but its frankly helped by the fact that I can get cleaned16 bit Imacon scans from 6x6 for the equivalent of $10/11 USA. But the real tester to me is the 36" sq. prints I've got here from 6x6 originals, via drum scan and Chromira, that are viewable from arms length. I couldn't get that with analogue printing from transparencies, Ilfochrome or otherwise. And I wouldn't get it unless all the elements - the original transparency, the scanning method and technique, and the printing machine and its management , were working well.</p>

<blockquote>

<p> </p>

</blockquote>

<p>"However, as the price of high-end DSLR going down, I have the intention to buy one like Philips's to replace my MF, not only benefits of ease of use, save processing time and cost"</p>

<p>When you see how much time and money people "invest" every couple of years into the latest RAW processing software, you might reconsider which is the most cost/time effective between MF and digital! As someone who uses MF and has shot consumer DSLR for years now, I can say that neither is easier to use than the other. With MF you even have some usability advantages, like not having to stand gawping at the back of your camera while things are happening in front of you for a start. Its a common myth to say that digital is easier to use. Its easier to use carelessley, and easier to produce hundreds of frames of crap, but not easier to produce genuinely good results.</p>

<p>I shoot a 5D mk1 and Mamiya RZ67 and scan with a coolscan 9000 and have printed from both. I find prints over 40x30cm clearly look better with MF. When you have the time and image quality is a top issue I think MF is a good choice. If large size is not a factor then 35mm is still worth shooting, and the look feel and quality is also tops at smaller sizes. But ISO on digital cameras adds a factor to speed and low light situations that is worth having.</p>

<p>I measured the resolution given by my Epson V700 flatbed scanner, it is about 45 line-pairs per millimeter.The 645 negatives are 42 mm high so comparing with a full frame DSLR 24 mm high this corresponds to 45x(42/24) = 78 lppm for full frame DSLR.<br>

This is the same as quoted above for the 5D11 so on paper at least it appears that even the more modestly priced V700 flatbed scanner working on 645 film should compare with a full frame DSLR.<br>

I never actually did this comparison so there may be other factors I missed.</p>

<p>I find that 645 is very similar to the 5DII here are two samples</p><div></div>

<p>Here is the m645 / Nikon 9000 - this is actually scanned at 2800dpi but 4000 dpi makes little diference</p><div></div>

<p>Now crops from GX680 and 5DII - same scene</p><div></div>

<p>5 DII for reference</p><div></div>

<p>Remenber these are very large enlargements - the GX680 is sized for 40"x30" and on my screen shows at a size that would be 100" by 72" - ie a 50 square foot print.</p>

<p>Philip, one of the reasons we did our tests with actual prints between a Mamiya 645 and D3 was because viewing screen resolution is a futile exercise when your intent is to make a print. A monitor is never going to look the same as a print.</p>

<p><em>"A monitor is never going to look the same as a print"</em> <br /> I fully agree and this thread obviously is not an apples to apples comparison. In the end however, the common element here is the post processing to get to that printed final piece. To answer the question of WS HO, there are many discussions in the Digital Darkroom section of PN</p>

<p>The limiting factor is the scanner.</p>

<p>Inherently, film has a wider color gamut and sharper resolution than a FF sensor. The color advantage is obvious when you're doing a wet-print.</p>

<p>I've only scanned on my Epson V700 at home and the Nikon 9000 at school. The V700 doesn't seem to match digital, although I've never tried a wet-scan. On the 9000 film seems to be the winner. I'm guessing that a Hasselblad scanner blows the digital away.</p>

<p>The main advantage with slide film is color correction. It took me a long time and a lot of different approaches to get a good color negative scan, but slide film requires little more than a simple color-cast correction, and usually not much of that either.</p>

<p>For those who are curious, my current negative workflow is:<br>

VueScan to produce a raw, unconverted scan.<br>

Assign the generic icc profile that came with my scanner to the scan.<br>

Convert to positive using ColorNeg.<br>

Convert to Lab and adjust luminance curve.</p>