scanned film vs digital

[Landrum Kelly]

<p>Yep, you guys were right in guessing that I have the 3800. I can hardly wait. I will have to buy some Velvet Fine Art paper, though.</p>

<p>Thanks to Dave, Les, Mauro, and everybody.</p>

<p>--Lannie</p>

[phil_derosier]

<p>Of course there's a difference ... film will be gone in 5 years, or less (unless you live in a third-world country village, whereby your film is processed next to that vat of questionable or illegal substances)</p>

[Dave Luttmann]

<p>Thanks for that highly informative post Phil. I take it you're a new DSLR user.</p>

[don_e]

<p>"(unless you live in a third-world country village, whereby your film is processed next to that vat of questionable or illegal substances)"</p>

<p>Um...you wouldn't have the name of that third world country village...?</p>

<p> </p>

[jeff_z.]

<p>Dave L, One of your recent comments got my attention, and I wondered if you or anyone else with similar experience could address my concern. About the 9000 you said: "..The Nikon 9000 is really capable of about 3800ppi, which is normally enough for most applications. Will you notice a difference? Yes....even at 16x20 from 120 film, the Nikon will be more detailed with less USM and have better shadow detail and tonality.."</p>

<p>Well, I have the older and lower spec'd., 35mm only, Nikon LS 40 (Coolscan IV). I love film and want to definitely stay with it, and careful technique with 35mm fulfills my individual needs. I've been making color prints from these scans for a few years now, and have sold prints made this way several times. Would the 9000 give results meaningfully better in the ways you describe (or otherwise)? It's probably important that I mention that I don't think that I have any desire to make very large prints, my absolute limit being about 14 to 15 inches on the long end from an uncropped 35mm original. I don't have access to a direct comparison, and wondered what you thought. Thanks.</p>

[Dave Luttmann]

<p>Jeff,</p>

<p>In terms of resolution, you probably won't see much int eh way of difference. If you do, it would be very slight. Where you will get a difference is in Dmax and tonal smoothness. The 9000, as well as the 5000 can dig deeper into the shadows. As well, the extra bit depth reduces the chances of posterization when making large tonal adjustments. If you're sticking with 35mm, get the 5000 insetad of the 9000.</p>

[Landrum Kelly]

<p>Citing experts on film-digital comparisons can be a rather treacherous undertaking. I went to the Roger N. Clark site cited above, and a <em>prima facie</em> interpretation of the data would seem to indicate the superiority of digital at its present level of development. If, however, one reads closely and <em>to the end</em> of the following links, one finds that even with regard to resolution, digital AT BEST matches 35mm film at about 15 to 16 megapixels--and again that is only with regard to resolution.</p>

<p>http://www.clarkvision.com/imagedetail/film.vs.6mpxl.digital.html</p>

<p>http://www.clarkvision.com/imagedetail/film.vs.digital.1.html</p>

<p>One may make whatever one wants of the results of the tests that Clark has carried out, but it seems pretty clear to me that film not only does quite well, but increasingly pulls away on all measures as one goes beyond 35mm film. For example, at first glance one of the above links seems to indicate that Fuji Velvia 50 seems to be on the level of equivalence of a 6MP sensor, but, if one reads to the end, the results indicate that a match in resolution occurs more nearly around 15MP--and again that is only with regard to resolution.</p>

<p>As one moves up to 6 cm x4.5 cm, the digital equivalence (for resolution only) is about 30MP, according to Clark, and roughly extrapolating to 6 cm x 6 cm one finds the digital equivalence to be about 40MP. At 4 x 5 inches there is, of course, no comparison.</p>

<p>While it is true that digital has made great strides and will no doubt continue to do so, one does have to ask whether a full-frame (24 mm x 36 mm) sensor can ever go much beyond 35 mm film, no matter how many pixels one tries to pack onto the sensor. </p>

<p>I did not casually plunk down $2K for the Nikon Coolscan 9000 today. I have been following these discussions for some years, and my own inquiries indicate to me that starting the switch to medium format over the last year was a lot more promising than continuing to watch the megapixel race unfold. I have no intention of getting rid of my 5D or the 1Ds Mark II, but for me the race ends there, at least for now. I will use digital at that level at current prices, but no further. New developments could change that, but right now the kinds of digital sensors that could match what I expect to be able to achieve with medium format film would cost me over $30,000, and I believe that to be a conservative estimate when I look at the cost of Phase One digital backs, for example.</p>

<p>So, whether I am right or wrong in my own decision today, it is a decision that I have been weighing for a long time, and I feel comfortable with it.</p>

<p>For me the practical corollary to all this is that, if one must win the megapixel race, one would more likely do it with scanned film. When one considers other factors besides resolution, film seems to be me to be the clear winner, except with regard to ease of use and processing time--but I know enough about time management to know that, as one becomes accustomed to anything, one can find ways to better manage the time issues.</p>

<p>Thanks for bearing with me today as I have thought aloud on this forum. I think that my range of both doubts and convictions has been pretty obvious. I claim no expertise, simply patient and reasoned inquiry. I will use both technologies as appropriate, but I feel very good about this decision, even if it means that I might have to unload an EF lens or two to break even today. The first thing to go will be the EF 24-105 IS, which to me is one of the world's most useless lenses for other than walk-around "utility outfielder" purposes.</p>

<p>I look forward to getting the best out of film for as long as it is there to be gotten, or until I die, whichever comes first.</p>

<p>--Lannie (aka John Henry)</p>

[bogdan_nicolescu]

<p>Here are couples of scans made with Nikon Coolscan 9000 ED using maximum optical resolution. No unmask or sharpening applied, contrast/brightness adjustements only.<br>

First shot was taken with Mamiya 7 II, 80mm lens, Kodak Portra NC, hand held in pretty low light. The crop on the right is from full size original image. Original dimensions are 36.7x29.9 inchs at 300DPI printing resolution.</p><div></div>

[bogdan_nicolescu]

<p>Second shot is a scan of 35mm full frame with Nikon Coolscan 9000 ED at full optical resolution. Shot was taken with Nikon F6, 35mm F/2 ZF Zeiss lens on Kodak E100GX slide, hand held.Original size is 18.8x12.5 inches at 300DPI printing resolution.</p><div></div>

[Landrum Kelly]

<p>Holy cow, Bogdan! Those are impressive. I'm feeling better and better as the evening progresses.</p>

<p>--Lannie</p>

[andrew_gale]

<p>Yes, I shoot almost entirely in film, but i still support digital as being the best way to obtain digital files. Unless you shoot medium or large format, you wont see much difference between high end cameras (D3 and D3X) and film in terms of quality.<br>

When I want a digital file of one of my photos, I make a print, then scan that.</p>

 

[bogdan_nicolescu]

<p>Some people might argue that the scans are not sharp enough...well... they might be right but I am perfectly happy with the output I get from both 35mm and MF formats. I still need to see a good example of what a high end digi camera can do.<br>

From my humble experience I can say that scanning films is indeed not a easy thing at all, it's time consuming and might be a very tedious job when the volume of work is high... I can also say that shooting film and digitizing the negs is not cheap at all but somehow I found interesting that the actual costs are not coming from the consumables but rather the original upfront investment in gear, scanners and computer power required...</p>

[danielleetaylor]

<p><em>Daniel, it has nothing to do with a sacred cow....it has everything to do with incompetent testing.</em> <br /> <br /> The author of Imatest and a NASA planetary imaging expert have incompetent testing? I suppose that's possible. Where can I find your published, peer reviewed articles and tests on the matter? You've got your work cut out for you if you're going to demonstrate that their sites are flat out wrong. I recommend a lengthy peer review before going public with your articles and tests.<br /> <br /> <em>I think we've all seen the Clarkvision, Koren, Luminous Landscape, Sphoto nonsense posted a million times by web-tards who can't think for themselves.</em> <br /> <br /> And I think we've all heard web-tards defend their sacred cows with nothing but empty words a million times.<br /> <br /> <em>I have a Canon D30. I used to own an Imacon 343. Funny, when I scanned Provia 100 and compared it to the D30 at 8x10, not a single person viewing the prints has EVER said the D30 was more detailed</em> .<br /> <br /> Reichmann displayed his prints in his gallery. Many people publicly agreed with him after reviewing those prints. Obviously there's a difference between your experiences. Perhaps subject matter, lenses, technique, who knows. Email him and hash it out. Actually you probably won't get a response because the D30 has been out of production for...what...7 years now? And he's probably busy working on his next gallery show, planning his next expedition, or preparing his next article or portfolio for publishing. In short, I doubt he cares.<br /> <br /> <em>As to the other sites quoted, a laughed at the "razor sharp" 20D at 13x19. I don't think manu people would find the print meets fine art works criteria of "razor sharp" with a 180dpi print.</em> <br /> <br /> In my experience, for most subject matter a 20D is razor sharp at 13x19. Not all subject matter. If the subject has really fine detail that makes up a large portion of the image, things like trees and grass in a landscape, then it's not going to come off as razor sharp. But then neither will 35mm film quite frankly. Both will come off as perfectly acceptable to viewers, but neither will look as good as a MF scan, or a current full frame DSLR, given that kind of subject matter.<br /> <br /> Many other subjects will look great however. Portraits, for example, can go even larger from a 20D and still look razor sharp and detailed. Subject is huge in these kinds of comparisons.<br /> <br /> <em>The main problem Daniel is that over the years, these sites have even contradicted themselves.</em> <br /> <br /> I don't see any glaring contradictions at Koren's, Clark's, or Sphoto's sites. If Reichmann refined his evaluation, perhaps after shooting a wider subject range, then so be it. He is only human and entitled to refine his position from time to time. So are the other sites for that matter.<br /> <br /> <em>The question has nothing to do with the resume of the owners of the sites you quote.</em> <br /> <br /> I'm not implying that you must agree with them because of their resumes. But by God, show a little respect. Name calling is something children do. And if you're going to pronounce to the world that they're misguided fools, you should have something to back up the claim. Clark has mathematical formulas accurately predicting the impact of noise on perceived image quality, derived from the same theories that guide the communications industry for crying out loud. Do you honestly think that calling him incompetent, with nothing to back it up, makes him look bad and you good?<br /> <br /> <em>The question is why, so many other photographers when viewing prints can't seem to replicate what these people come up with.</em> <br /> <br /> Who are these other people? Where is their opinion documented? No fake statistics please. If you're going to appeal to some mass majority, please at least demonstrate that said majority exists.<br /> <br /> <em>Oh, and by the way....you ever notice how it's the same 3-4 sites constantly quoted by the anti-film brigade?</em> <br /> <br /> What's an anti-film brigade? I didn't know such a thing existed. Is it a Canadian thing because I've never heard of one here in the states?<br /> <br /> And have you ever noticed that two of the sites have some of the most well executed and well documented tests and theories on the subject? Koren and Clark could publish in journals. They're that good because, well, they publish in journals for a living. Their sites are some of the most cited and referenced web sites on imaging technology period for that reason. Once again, Koren is the creator of Imatest, and Clark is a NASA expert on planetary imaging. If that doesn't humble you and make you second guess any disagreements you have with them, then you are far too arrogant a human being. They're not perfect, they can be wrong, but I wouldn't be too sure of myself against them unless I had my tests and math well in order if I were you.<br /> <br /> <em>Funny, film users point out hundreds of sites and samples available proving their point,</em> <br /> <br /> Hundreds? If I Google phrases like "film vs digital", "5D vs film", "1Ds vs 35mm", etc I find articles and tests which all pretty much come to the same conclusions, give or take a few degrees of difference. It's interesting to note that far more has been written about full frame versus medium format than 35mm, which pretty much demonstrates that people think the latter debate is done, that is that digital is comparable or even better. Where are these hundreds of sites proving Clark, Koren, Sphoto, Reichmann so horribly wrong? Les is a die hard film fan, but even his maps show that around 12 MP (D2X frame) resolution comes down to nitpicking while digital exhibits better tonality and noise versus 35mm.<br /> <br /> To me the remaining appeal of film is the price/performance of larger formats and the unique looks or palettes of certain films. I won't go as far as some to say that those looks are impossible to replicate digitally, but they're difficult to replicate consistently unless you shoot a reference frame of the same scene on the film. And if that's the case, you don't need the digital file. If someone likes the feel of Velvia, shoot Velvia. It's a wonderful film. If someone picks up a used MF or 4x5 rig, by all means, please, load it up and shoot! They are capable of great results. There is no "anti-film brigade". Personally I just don't shoot that much color film any more, but I'm not "anti" anything.<br /> <br /> I said it earlier and I'll say it again: if digital didn't deliver the goods the majority of pros would not have switched to it. Film delivered the goods for years. So pick one and go shoot something.</p>

[danielleetaylor]

<p>Landrum,<br /> <br /> <em>I also know that Daniel also appreciates film, too, and so I regret seeing this thread become unnecessarily polarized, as tends to happen on this issue.</em> <br /> <br /> I still shoot, develop, and print B&W at the local fine arts center...which has a killer rate on use. While I feel like my digital conversions are roughly equal now, I enjoy it and it keeps my eye calibrated, if you will, for when I do digital conversions. I also love infrared film.</p>

<p>But the anti-film brigade is sending me ads with special discount rates...I don't know, I'm tempted to join. They have pitch fork and torch meetings every month, and once a year they have a film burning. It sounds like fun, don't you think? ;-)</p>

[Landrum Kelly]

<p><br /> "To me the remaining appeal of film is the price/performance of larger formats and the unique looks or palettes of certain films. I won't go as far as some to say that those looks are impossible to replicate digitally, but they're difficult to replicate consistently unless you shoot a reference frame of the same scene on the film."</p>

<p>I think that you nailed it right there, Daniel, and I have appreciated your links. I also know that you shoot both and know the strengths and limitations of both better than I.</p>

<p>As for the pros, well, time is money. What else can I say? Digital is faster and sometimes better, depending on what one is trying to achieve.</p>

<p>Therefore, "So pick one and go shoot something"? Nah. We can shoot both, as they are appropriate. I've been shooting both since 2002, and this step up for me to medium format is probably pretty much the demarcaction line. That is, if things go well, I will probably shoot mostly digital when I want the convenience of 35mm, and I hope that I will have the patience to shoot MF most of the rest of the time. </p>

<p>What I don't need is another piece of equipment that is just gathering dust, but I hope that the challenge of good scanning will keep that from happening with my new purchase. I don't think that I will be sitting around saying, "Gosh, for only $3000, I could have had a 5D II with a battery grip." The marginal utility over my existing 5D and IDs II is simply not there.</p>

<p>Thanks for putting me onto the Clark site.</p>

<p>---Lannie</p>

[jeff_z.]

<p>Dave,</p>

<p>Thanks, that's just what I was looking for. It's funny though, it'll be hard to part with this LS-40 until I can actually visit a store and perhaps make a comparison scan on the 5000, and see for myself, as I've had such good results, and viewer feedback, and some sales:) </p>

<p>I really appreciate your taking the time to answer my question. Jeff </p>

[Landrum Kelly]

<p>Daniel, I ride a 1982 Honda CB900C and I smoke Harleys, but I don't burn Harleys the way some of the "Born to Be Loud" crowd burn Japanese bikes. I just like to ride, and I do.</p>

<p>As for photography, I just like to shoot, and I do. I don't join brigades. I pretty much go it alone, and I like the prospect of having two ponies.</p>

<p>--Lannie</p>

[danielleetaylor]

<p>Landrum,</p>

<p><em>Thanks for putting me onto the Clark site.</em></p>

<p>I noticed you actually read it and realized that he ranked film pretty high in some respects :-)</p>

<p><em>Daniel, I ride a 1982 Honda CB900C and I smoke Harleys, but I don't burn Harleys the way some of the "Born to Be Loud" crowd burn Japanese bikes. I just like to ride, and I do.</em></p>

<p>It's human nature. I have two uncles who will endlessly debate Ford versus Chevy. Who knows why...</p>

[Dave Luttmann]

<p>I did notice one thing Les, unlike Daniel, I use my eyes. Viewing Reichman's prints myself...something Daniel hasn't done...I can say beyond a shadow of doubt that 3mp didn't match Provia. Daniel of course still quotes the site.</p>

<p>One thing you'll notice, unlike those of us who have posted samples, Daniel doesn't seem to be able post any comparisons with high end scanners.....just links to long refuted sites. Oh well.</p>

[rjacksonphoto]

<p>With all due respect to The Da Vinci Code, strictly speaking, scotoma refers to a blind spot. Which, come to think of it, might actually be quite appropriate for this entire thread. ;-)</p>

[jack_welsh]

<p>"I said it earlier and I'll say it again: if digital didn't deliver the goods the majority of pros would not have switched to it. Film delivered the goods for years. So pick one and go shoot something."<br>

Daniel, at first the pros switched to digital not because it was better than film in photo quality. But, because it had a faster turn around time and their competition started using it for the same reason.It was too expensive to buy a digital back for their MF cameras. Even though the digital cameras were not as good as even a 35 film camera! When, I was buying the photo mags several years ago. They finally had articles where a $13,000 Kodak and a $8,000 Canon finally was equal to or better than 35. Big deal!What isn't?</p>

[george_peng1]

<p>

 

<p>I'm not sure I understand the point here. I think the easiest way to think about them is as different tools, and hence different tools for different tasks.</p>

<p>As for resolution, I don't think its much of a contest. My scanner scans at 7200 dpi, making a full frame negative scan at about 69mp. Obviously no 35mm DSLR is going to do that anytime in the next few years. It's well-documented that film dmax is 4-8 stops wider than digital, so that's another issue.</p>

<p>The functional question is whether you can see the difference and which fits into your workflow needs. Obviously, digital gives you great instant feedback on composition and exposure. That's obviously invaluable, to the extent the results are adequate to the purpose.</p>

<p>That said, getting a D90 DSLR has made me appreciate analog more, especially my Contax G1 with Zeiss lenses. It's going to be hard to convince me that a DSLR can match up with those lenses with Astia or Provia. But once again, its a matter of where you want to be in terms of trading off IQ vs. convenience.</p>

<p>But as a theoretical aside, while scans are arguably a 2nd generation product, I think its incorrect to say that a DSLR image is 1st generation, simply because the DSLR sensors are digital and therefore not only subject to data depth but also Bayer interpolation, AA filters, etc., which is therefore inherently reductive when compared to any analog capture. The same issue arises with respect to any digital media, be it music, film, etc. - mp3's are worse than CDs, which are worse than vinyl. The quality of the digital copy is inherently limited by the sampling rate, so unless your sampling rate is infinite, then you are only making an approximation of the analog source.</p>

<p>But of course, the question is whether you can tell and whether you care. And given the profusion of iPods, etc., its clear the market has made its decision, which is to say that it works for most people, but there will always be market segments who are willing to trade convenience for quality.</p>

 

</p>

[danielleetaylor]

<p><em>Well below is the the same 35mm Fuji RVP map . . . but this time scanned at 8,000dpi on an Imacon 949.<br /> 8000dpi of Fuji RVP 35mm - > Nikon D2X - > Click thumbnail for full res</em> <br /> <br /> Thank you for giving me the opportunity to make my point. Attached is the Velvia scan next to the D2X crop resized. The D2X crop is sharper, has better tonality, and is cleaner. It's rougher on the edges, but then the Velvia is rather blurred on the edges. Neither out resolved the other, it's just blur versus a certain roughness that comes from scaling up. At 50% Velvia still looks blurred, but the D2X crop looks crisp and very well defined.<br /> <br /> <em>I know it goes without saying but just to point out a couple of particularly obvious differences, please look at Libreville and the formation of the star symbols. Do you believe the difference between these two are "nitpicking "?<br /> </em><br>

Yes. A fight between a group of PC and Linux nerds would involve less nitpicking.</p><div></div>

[danielleetaylor]

<p><em>Viewing Reichman's prints myself...something Daniel hasn't done...I can say beyond a shadow of doubt that 3mp didn't match Provia.</em></p>

<p>I spent the majority of my post defending Clark and Koren, and Reichmann's site and work in overall terms. I don't think much either way of Reichmann's 7 year old article on a 7 year old camera, and I didn't really address it as a point other than to tell you to hash it out with him. Yet you're stuck on it.</p>

<p>You claim to have gone to his gallery and viewed the prints directly. I'm asking honestly: did you have a fight with him over it? Is there some bad blood between you two? There's too much personal obsession from you over this one article from this one man. I wrote pretty much about other people and other things, and yet you're stuck on this. There's something more to the story and it is coloring your ability to discuss this issue in a rational, adult manner. I'm curious to know the source of the obsession.</p>

[danielleetaylor]

<p><em>One thing you'll notice, unlike those of us who have posted samples, Daniel doesn't seem to be able post any comparisons with high end scanners</em></p>

<p>You've posted no samples or comparisons, and there's no link in your profile to any site. So this was a pretty stupid dig on your part.</p>

<p>But if you have work comparable to Clark or Koren, I would love to read it. Seriously, post a link, I'll read every paper and look at every scan.</p>