Do you scan?

[steve_rasmussen]

<p>I shoot 35mm, 6x7 and 4x5 slide film but only have my 6x7 slides scanned (subcontracted). The reason is twofold. My 35mm and 4x5 work is done for fun, while my 6x7 work is for production. The other reason is that most of my customers no longer accept film as a submission media. I only have my best work scanned, as doing everything gets too expensive. The scanner is a Heidelberg Topaz</p>

[lorriman]

<p>Anyone tried 'scanning' negs with a DSLR? This is how I do my 35mm film. </p>

<p>I would suspect that shallow DOF portraiture, my own concern, would not be so high in resolution that a 16MP DSLR couldn't do justice to a 645 negative. </p>

[jens_g.r._benthien]

<p>A 6x4.5 negative has an area of 56x41 mm or 2.2x1.6 inch. Scanned at 4.000 ppi this results in an image of 8.818 x 6.456 pixels which is 56.93 MegaPixels @ 48bit 'off sensor'.</p>

<p>Your 16 MP are just 28% of what your 6x4.5 negative delivers.</p>

[lorriman]

<p>The resolution of a scanner isn't the resolution of the negative. A 35mm negative gets to 10x8"which even my 6MP Fuji F30 can print to easily. So I think your figures are a bit off, gens.</p>

<p>35mm is about 8-10MP at the very most (colour; B&W may be 25MP from reports). With an open aperture and 400 ISO film I would expect perhaps 4-8 MP from MF. That would be within the realm of a 16MP DSLR copying process. The latter is guess, of course. That's why I'm asking if anyone has tried.</p>

[q.g._de_bakker]

4000 dpi scanner resolution will not surpass image resolution in a negative. Those calculations Jens gave are spot on: about 57-58 MP (depending on the real size of a 6x4.5 negative) of true image resolution in a 6x4.5 negative.<br>I know, because i have tried (do so on a regular basis).<br><br>A 6 MP image will be just about enough for a digital (!) print in the region of 8x10". Nowhere as good as a wet print from even a 35 mm negative.

[lorriman]

<p>Seriously, you two need to get a reality check. I'm keen on film but not for the crappy resolution. This is taking film nuttiness to another level; not one included in Hawkings' 11 dimension either.<br>

In the meantime please let others respond to my post with their 'alternative' views. And thanks for trying.</p>

[philip_wilson]

Greg I can assur

E you than my 645 scans on a Nikon 9000 are about the same resolution as my 5DII (22MP) and that my Fuji GX680

clearly out resolves the Canon. This is true even with colour slide film. You can search the site to seen the crops from

these images if you doubt this. Obviously the workflow from my Canons is much quicker and simpler than scanning. My

own subjective assessment is that 645 is about the same as 20 -25 MP digital from a full frame camera and that 6x8 is

probably about 40 MP equivalent. Of course with faster film or cheaper scanners the results will be worse - my examples

above are for my Nikon 9000 using a carefully prepared slide and a very high resolution scanning setting. They are also

based on Velvia 50 or 100. A cheap 400 ISO print film in a flatbed scanner will be a lot worse.

[jens_g.r._benthien]

<p>Greg, maybe you should read <em><strong><a href="https://toyotadesigner.wordpress.com/category/photography/resolution-of-film/">this</a></strong></em> before you'll waste your time.</p>

<p>Scroll down to part II and part I if you are good in mathematics.</p>

[q.g._de_bakker]

Or better forget about mathematics (numbers can mean anything) and take the word of people who actually know what you can and cannot get scanning film.<br>In the past i have posted a 4000 ppi sample scan on PNET. It showed that at single pixel level, in parts of the image that are rather featureless you mostly see grain and noise. In parts that do show bits of the subject with a lot of fine detail, the pattern of grain and noise is considerably different, i.e. the detail present is captured. At pixel level, A higher resolution scan may still not just be enlarging grain. That is in complete agreement too with the figures Zeiss once published about film resolution in everyday, i.e. non-laboratory test, settings.<br><br>If anyone needs a reality check, Greg, it's someone who thinks film has crappy resolution and favours digital capture's even crappier resolution because of that.

[ondebanks]

<p>Jens, all your (lovely!) examples are on fine-grained Provia 100F.</p>

<p>IIRC, Q.G.'s sample image in the previous thread was also on a slow film.</p>

<p>But Greg is talking about ISO 400 colour film. That distinction has to be addressed.</p>

[mwmcbroom]

<p>Yes, I scan in almost all of my emulsion-based photos. Not for archive purposes, but for greater ease of electronic manipulation and transmission. But like so many others, I've been chasing the holy grail of maximum information extraction from my slides and negatives, and its been a tough and often frustrating road.</p>

<p>I employ two main means of image "scanning." I have an Epson 4990, which I use for medium format and non-critical 35mm. I also have a slide duplication rig I've set up for use with my DSLR. I find that it is not too difficult to also dupe negatives with this rig. Photoshop's negative conversion process is relatively painless.</p>

<p>I find the Epson's resolution to be adequate for medium format because of the size of the image. Even though Epson claims a max resolution of 4800 ppi, its actual resolution is more like 2000 ppi, and while I consider this to be insufficient for producing critically sharp 35mm scans, it is plenty good enough for medium format scans and enlargements up to 20" or greater.</p>

<p>My slide duplication setup is a great way to "scan." It's fast, the scans are very crisp, with accurate color, and really, the list goes on. If I were to use a Canon 5D Mk II as the camera in my dupe rig, it would be capable of delivering images just shy of 4000 ppi. And those are actual numbers, too. Not the fudged numbers that most scanner makers foist upon an unsuspecting public. Nikon Coolscan specs for less money, plus a great DSLR is included at no extra charge!</p>

<p>Unfortunately, I'm still saving up for the 5D Mk II, so I have to make do with a DSLR that delivers "only" about 2700 ppi, which is still quite a bit better than any flatbed scanner I'm aware of. I can say this regarding the finer grain films I've scanned, such as Kodachrome 25, Ektar, and Fuji Velvia -- 2700 ppi is insufficient for capturing all the detail available in a 35mm image. A simple visual examination with a good loupe is all that I require to confirm this.</p>

<p> </p>

[ondebanks]

<p>In fact, I just noticed this...</p>

<p>Greg:</p>

<blockquote>

<p>35mm is about 8-10MP at the very most (colour</p>

</blockquote>

<p>Philip:</p>

<blockquote>

<p>My own subjective assessment is that 645 is about the same as 20 -25 MP digital from a full frame camera</p>

</blockquote>

<p>645 is 2.7x the area of 35mm film, so Philip's result (with a 4000 dpi Nikon 9000 scanner) scales to 7.4 - 9.3 MP for 35mm. Exactly in agreement with Greg.</p>

 

[lorriman]

<blockquote>

<p>jens:"<em>Greg, maybe you should read <strong><a href="https://toyotadesigner.wordpress.com/category/photography/resolution-of-film/" rel="nofollow" target="_blank">this</a></strong> before you'll waste your time.</em><br>

<em>Scroll down to part II and part I if you are good in mathematics.</em>"</p>

<p>jens, man, you have got to be kidding,and definately kidding yourself. This is like the issue when people say that 320Kbps mp3s (or FLAC) sound better than 192Kbps, when in double blind testing no one can tell the difference. Look at the crop: it isn't anywhere near digital. Just doing a simple resize (Lanczos) requires taking it down to 17% of the area (186 pixels on the horizontal), ie. like going from 10 MP to 1.7MP, to get a visual equivalent to anti-alias filter digital (ie. detail but slightly unsharp). Here's the resize:</p>

<p><a href=" src="http://farm9.staticflickr.com/8174/8026458174_6f31208f94_m.jpg" alt="" width="186" height="127" /></a><br>

In fact I think a little more reduction is needed.</p>

<p>Now, sure, the image might technically even be very high resolution as you claim, but that's comparing film grains and not colour resolution. The maths is meaningless because it doesn't allow for the difference of the mediums. You need huge amount of grains to make the equivalent of a single colour pixel, as is obvious from your examples. You are only going to get your kind of resolution from meaningless test charts. The other examples are the same. Read Luminous Landscape or some other more authoratitive site if you want the real low-down on this issue. In any case your example pics speak for themselves.</p>

<p>Roughly speaking 35mm film, on a good day, is equivalent to 8 to 10MP digital in REAL LIFE; resolution test charts notwithstanding. </p>

 

</blockquote>

[lorriman]

<p>oops, put my whole post in quote colours there. </p>

<p>Thanks for the replies from Philip, Ray and Michael. </p>

[starshooter]

<p>I have 35mm and 120 sized images that I shot in the 1960s and 1970s of Elvis, the Beatles, Lucille Ball and so on. I scan my negs with an Epson V500 Photo scammer. I had some 11x14 and 11x16 photos-on-canvas made and the lady who made the prints said the scanned images could be blown up four times larger than those I had made.<br>

I currently shoot 35mm and 120 film and scan the ones I want to enlarge.</p>

[q.g._de_bakker]

Ray,<br><br>My sample was on either (i forget) Fuji 100 or (more likely) my 'stock' film: Portra 160 NC.<br>Not particularly fast. But not a slow film either. A good film, Portra, yes. But nothing special, neither one.<br><br>But let's not get this thread off-track. The nonsense about films 'MPs' has been identified and put straight (for those wondering what nonsense that is: it's the content of Greg's posts). So a return to the original scanning question?

[lorriman]

<blockquote>

<p>Bakker: "But let's not get this thread off-track. The nonsense about films 'MPs' has been identified and put straight (for those wondering what nonsense that is: it's the content of Greg's posts). So a return to the original scanning question?"<br>

</p>

</blockquote>

<p>Guffaw!!!</p>

<p>You aren't fooling anyone. And the attempt to avoid the content of my reply is just inane. </p>

<p>Somehow you've convinced yourself that film is unbelievably superior to digital in resolution despite the evidence of both our and your eyes. Now that's really quite bizarre. The demonstration I gave makes the mistake obvious. Perhaps to save face or something you are continuing this fiction. I think you should apologise and bail out. </p>

<p>Film is great for DR, which is what I need it for. But digital is not far catching up with that to. 5 years from now and I doubt I'll be using film.</p>

[q.g._de_bakker]

Greg,<br><br>I do know what i am talking about. Do you? No, you don't. Obviously you don't.<br>Your choice of words betray the nature of your knowledge. <i>"unbelievably superior"</i>? A rhetoric hyperbole? Nah... Lunacy of the adolescent "we'll show you who is master" kind.<br>I have tried and tested, and the results are the results. Your demonstration demonstrates what? What is it that <i>you</i> have demonstrated?<br>You have demonstrated a considerable degree of ignorance by what you said about grain vs pixels. <i>"huge amount of grains to make the equivalent of a single colour pixel"</i>...<br>So if anything, Greg, the only thing that's obvious is that you are championing digital in a digital vs film lunacy, without knowing much about either.<br>So if we're inviting people to do something, i would invite you to first free your mind of preconceptions, learn about both media, then try to give an informed 'demonstration' of what you found out. (But please not here: we don't need another silly (as in: look at the nonsense being spouted) digital-is-superior-to-film 'discussion'. This thread had been kept going by people who do know the merits of both, until Maris tried to derail it, and until you came up with this misinformed nonsensical <i>"you have got to be kidding"</i> idiocy.

[lorriman]

<p>Ho ho. I suppose there really is nothing to be done. These guys are totally bananas. Indeed they are quite difficult to tell apart. :) </p>

<p>In the meantime for the sane, here is a genuinely more authoritative opinion, rather than some random and bogus blog :</p>

<p><a href="http://www.luminous-landscape.com/essays/clumps.shtml">http://www.luminous-landscape.com/essays/clumps.shtml</a></p>

<p>And apart from that just use your eyes. Take those example crops given by our delusional buddies, resize them down to the subjective image quality of a decent digital image at 100% crop. You'll probably also hit around 17% of the image size. Take their crazy resolution figure of 58MP and multiply: 58*17%= 9.86MP, and that's being generous with the numbers. That's slap bang in the range of everyone elses experience and that of real authorities. 35mm film comes to about a practical equivalent of 8-10MP. </p>

<p>That means that 645, on a good day, is 24-30MP (3x) if we are generous with MF lens resolution which is said to be less than a typical 35mm lens. Wide open with 400ISO? Presumably a lot less. Even perhaps justifying copying with a 16MP DSLR.</p>

<p>Real world experiences would be good to hear of, of course.</p>

[q.g._de_bakker]

Yes, Greg, just use <i>your</i> eyes.<br>Good advice, that. Some people have. Some people rather read what their like-minded come up with on the internet (and then it, of course, must be true).<br>Well, now you introduce yet another of those silly believes people who do not use their eyes like to repeat: MF lens resolution... <i>"Said to be [...]"</i> Indeed. By people who do not use their eyes, and only know what they read on the internet.<br>I'm sorry Greg, but it's quite obvious that you really have no clue. Yet it's not difficult to find out how things are. But (and it should go without saying) that means going off the internet, and <i>"just use your eyes"</i>.<br>You ask for real world experiences? It didn't get through to you that you were given just that: real life experience.<br>But (<i>"you got to be kidding"</i>) you apparently rather believe what you read on the internet. Rather selectively too: only those bits that agree with your preconceptions. Well, Greg, just go and use <i>your</i> eyes.

[tom_bloomer1]

<p>According to the cited "authority" <br>

"- it takes a clump of between 30-40 grains of film to represent a full tonal range"<br>

If that kind of over simplfied math (emphasis on simple) gives you headache, take 3 Bayer asprins and interpolate them. <br>

. . . and mine goes to eleven.</p>

[jon_kobeck1]

I send it all out to Northcoast. The enhanced scans are marvelous and cheap. About $20 a roll with processing. It's just

not worth my time to try to do it myself, and I can not get the same quality with my Epson flatbed that they get. And they

just updated all their equipment

[mwmcbroom]

<p>Ray, regarding your post where you cite comments from Greg and Philip on perceived MP values of color film, I must say that these numbers do not agree with my own assessments. 8 to10 MP really is on the low side for a good color emulsion. I'd say 8 to 10 MP for maybe a decent modern ISO 400 color emulsion, yeah, I'd buy that. But definitely not for an emulsion like Kodachrome 25 or Fuji Velvia. I have a 10.1 MP DSLR that I routinely use for duping slides (the lens is a superbly sharp 55mm f/3.5 Micro-Nikkor) and I <em>cannot resolve the grain from Kodachrome 25 or Fuji Velvia.</em> I can't even really resolve the grain from K64 slides or modern emulsions such as Elite Chrome 100. But I <em>can</em> resolve grain from older emulsions from 20+ years ago, such as regular old Fujichrome 100 or Ektachrome 160. So, yeah, the old regular stuff is about 8-10 mp, but not the really fine stuff, no way.</p>

<p>Film, by its very nature, is not going to be nailed down with the precision that digital technologies are. It's got all that messy chaotic randomness going on that digital never has. Which is part of its charm, you ask me. Nonetheless, film's <em>theoretical</em> information content can be quite high. Consider this little bit of math that anybody with a calculator can follow: if we assume that a very fine grain film exists that can resolve information to the fineness of 100 line pair per millimeter (and this is a safe assumption because there are film emulsions out there with this sort of resolving power), then this is the equivalent of film containing 200 pixels of information per millimeter (one line pair being equivalent to two pixels, yes?). So, one 36mm x 24mm image taken with this film will deliver (36mm x200 lines/mm) x (24mm x 200 lines/mm) = 34,560,000 lines squared, or let's just call a "square line" a pixel, like the sort of square pixel you have on flat screen monitors and laptop screens. So, this means that an image on a fine grained 35mm film emulsion is capable of containing about 34.5 megapixels of information.</p>

<p>But wait -- when's the last time you've seen specs on a lens that has 100 lp/mm or better resolution? Typically the sharpest lenses usually can't handle more than about 65 lp/mm or so. So, let's use that number. (36mm x [65lp/mm x 2]) x (24m x [65lp/mm x 2]) = 14,601,600, call it 14.6 megapixels. Still not too shabby, eh?</p>

<p>Now granted, things tend to get rather messy in the real world, and we don't often see these sorts of results. But I think that it can be successfully argued that under certain ideal circumstances, even 35mm images can rival those of the best digital cameras. Math doesn't lie.</p>

<p> </p>

[djphoto]

<p>I scan 35mm film with a Minolta-Dimage 5400 unit, which gives me 13x19-inch scans that easily make sharp 20x30 prints. For medium format I use an Epson 4990 with the Epson software. The largest print I've made from that setup was a 40x60 from a 6x7 Fujichrome 100 transparency shot in a Pentax 6x7. I have been a full-time professional since 1972, so I'm not totally ignorant of this field.</p>

<p>This whole digital vs. scanned film thing which has been under such intense discussion became a non-issue for me in 2003 shortly after I acquired my very first digital camera. The gig was a bridal portrait, which I planned to shoot with my standard rig at the time, a Pentax 6x7. I didn’t yet know or trust my shiny new Canon 10D, but thought I would make a few shots with it for comparison purposes.</p>

<p>The 10D files showed promise, even though from only a six-megapixel camera, so I had 16x20 prints made of similar poses, one from a 10D jpeg and the other from a professionally scanned NPH negative. When the lab owner gave me the prints, he said "Hmmmmm." Another local pro, who worked with the Mamiya RZ system, came in as I was looking at the prints. He said "Hmmmmm." Then he said "Well, I was saving up for a trip to Europe, but it looks like I'm going to have to spend my money on something else."</p>

<p>I showed the prints to others of my fellow commercial photographers and several of the art directors I work with. Only one could tell which was which, and I later learned she recognized the digital file because it had more depth of field.</p>

<p>Some of you guys say you can tell the difference between film and digital. Maybe you can, but I can tell you that a bunch of experienced professional photographers and art directors in my city couldn’t. And neither can I, except that digital usually looks better.</p>

<p>I will concede one thing, though. Even though digital is better in almost every way, film was more fun. Or maybe I’m just waxing nostalgic.</p>

[q.g._de_bakker]

I'm sorry, Dave, but we have been told such stories ever since digital cameras were sold as "professional" tools, sporting 2 or even 3 MP sensors.<br>All we know from such stories (and many comparisons out there on the net) is that some people manage to make the results from film come out as bad as those of mediocre digital capture. It's not true: 6 MP capture can not even equal or better film at 6x8 prints. I know. It just is not so.