DSLR (Nikon/Canon) mounted on Hasselblad Bellows for macro reproductions of flat materials with V system zeiss lenses?

<p>I would like to ask for a little bit of wisdom guys as I find myself yet again in a pickle.<br>

<strong>I wish to reproduce film using a digital slr</strong> however I have been unable for the past month to find an affordable setup (body, lens, extension tubes) to do so that would give me the quality i require.</p>

<p>The problem is that I require very high pixel counts on the final composite image so with just extension tubes on a DSLR setup I can get at most 1.5:1 magnification so after all the math I would need at least 20 MP digital slr for my purposes. There are bellows for dslrs but it's yet another expense for a piece of kit that will not be used for any actual photography.<br>

So since I already own high quality zeiss lenses (50mm, 60mm, 80mm, 180mm) for my hasselblad V system and a bellows <strong>I would like to pick the lens that is sharpest out of my current set</strong> in the cropped area that would be picked up by the DSLR for the given task and <strong>shoot away on an old DSLR</strong> (that offers say 12MP - enough when shooting at 2:1 magnification for instance and just compositing more images together for the final product) <strong>mounted via a 3rd party watchyacallit on my old style hasselblad bellows</strong><br>

<strong>Are there any issues that come to mind with this idea?<br /><br /></strong>It is a pain in the ass to composite ~12 images into a single medium format "scan" but this is only for final printing mind you. For proofs and the web I would just composite out of 4 images. <br>

I could just get a Nikon D800 and a 1:1 lens and a single extension tube but that is 1500 pounds cost and for that money I could get an older digital back for the hasselblad over the next two years so not being a fan of these modern DSLRS and not needing any kind of automated features I just don't want to sink all that money and then see it depreciate in value as I use it only for reproductions.<br>

Would it be sensible sharpness wise to do 12 images at 2:1 magnification using the bellows and say the standard 80mm? I do not have the 135mm at the moment but I do have the 180mm so maybe the 180 would be even better? At 2:1 it is close to the subject enough that it is not a big deal but will I see any more sharpness from the 180 to warrant the extra distance (longer masking for straylight required) to the light table?<br />Will the 135 being optimised for proper macro work blow me away compared to an 80 on the bellows for this purpose?<br>

<strong>Being dedicated to the hasselblad system I do not want to split the budget in the middle</strong> for yet another system so if I can make this work I will just throw away the "just for a year or two" old dslr and replace it with a digital back eventually for composite reproductions. <br>

I have no idea how it will work sharpness and colour wise given the long bellows extension and the crop factor and etc. so I would dig it if you guys could tell me how it is out there in the digital world so that I can waste less money trying out the wrong bits and pieces. I'd much rather get a beautifull sinar setup for 8x10 for the cost of a dedicated modern dslr setup than a digital slr and avoid the inevitable hair loss when in 3 years I see people take 36 mp shots with their phones.<br>

16mp DSLR can be had for a couple hundred quid - also any recommendations on that front if my idea is in fact not that crazy? Canon or Nikon? Maybe another brand? If the hassy bellows idea is workable then colour reproduction and <strong>the possibility to see a live tethered view on the laptop</strong> takes precedence over pixel count.<br>

I would also be building my own rig for mounting of it all above the light table but have a large vintage berlebach that could also do the trick for the time being.<br />I mostly want to go this way because I already have everything that it would require apart from the digital body and the mounting converter to start experimenting right away and since film is waiting to be scanned time is pressing for me at the moment.<br>

Much high fives,<br />BP</p>

<p>The lenses you are thinking of using are not realy optimized for macro use. Slide copiers are very reasonably priced and usually use enlarging lenses. If the slide copier comes with the enlarging lens that's fine. If it doesn't then you can get one easily enough. You can then buy some memory cards and rent a new D810. A medium format DSLR with a sensor not much larger than 24X36mm will not give you much better results than the Nikon D810. When the improved version of the D810 comes out you can rent that model instead. Whether you have enough film to be scanned that a dedicated scanner is worth buying is another question. For now a slide copier using an enlarging lens and a rented Nikon D810 should give you adequately high quality results. </p>

<p>The problem is that I need 120 MP and up to make it worth my while to do it this way at all, otherwise it makes more sense to send it off for scanning on a drum scanner.<br>

The idea is to have a composite image so the area of the sensor is not important as that would be offset by the magnification and the amount of individual shots taken to arrive at a composite with enough resolution.<br>

I did inquire about renting gear but that is about 40 quid for a day so if I can pay 200 quid and have something that will work just as good but use my own lenses and have it for a few years then it is a lot more affordable.<br>

The problem is that with renting I would need to buy a cheap scanner for proofs and then rent and do my proper to be printed run in one day with the rented gear which is in terms of time more of it spent than making a more time consuming composite with an older dslr.<br>

<strong>Oh and just for clarification, I want to use a 35mm dslr with a mounting convertor on my regular medium format rig - not a medium format dslr.</strong> The crop would be big but that is not an issue as most lenses are sharpest in the middle so I would be shooting with the sharpest part of the glass anyway.</p>

<p>You should go over to the large format forum and take a look at the project they've been collectively running for quite some time now. The idea is a computer-operated system that uses a DSLR to scan large format film and stitch it. They're using a computer-driven gantry like a CNC router to do the shooting</p>

<p>There are a set of threads addressing each problem, lighting, camera, lens, etc. Start here:<br /> http://www.largeformatphotography.info/forum/showthread.php?87536-DSLR-Scanner-Light-Sources to see a list of the separate threads.<br /> The whole thing started here:<br /> http://www.largeformatphotography.info/forum/showthread.php?84769-Making-a-scanner-with-a-DSLR</p>

<p>I'm doing single shot scans, myself. My setup is here:<br /> <a href="

Except for the use of your Hasselblad lenses, this is what you want:

http://www.cambo.com/html/products_photo/set01/english/internet/item85.html

 

Keep the lens stationary and shift the camera that is mounted on the rear standard. Keeping the lens in a fixed position

and moving the camera instead eliminates parallax.

 

There are other ways to mount a DSLR on a monorail view camera as well, but the issue is that the depth of the mirror

box in the DSLR limits the shift range.

<p>Hey guys,<br>

I will have a look at that automated rig but just to clarify the hasselblad bellows are not intended for movements but for macro work. I do not need movements to reproduce flat materials at these magnification levels, there is no visible perspective to flatten out.</p>

<p>The reason I want to use my medium format rig to scan medium format film but with a 35mm dslr at the end of it rather than a dedicated dslr rig is to simply save money as I have no desire to shoot small frame digital. So I already know the dedicated rig solution works so I just want to figure out if there are any issues with just using the 35mm dslr body with my regular medium format system.</p>

<p>The problem is made worse by the fact that you can't just buy a digital camera without all the ruggedness and the multitude of completely pointless features like iso in the thousands, 10 frames per second or autofocus which I have absolutely no use for so I want to grab an older dslr for less money and make it work for me by using high magnifications. 12 shots stitched at 10 MP of useful information per shot is 120 mp end result which gets me in the ballpark of what I need for my prints. So a 12 mp small frame dslr would be enough in this case.<br>

Ofcourse I'd rather use a Nikon d800 and grab 4 shots instead of 12 but I just hate having to spend money on something I just will never, ever use and it will just sit there and deprecate in value on the shelf.<br />Therefore I want to figure out if the specific lenses for the specific setup I mentioned will work well enough before I actually commit to buying an older dslr because if there is some issue after the fact and I need to go dedicated dslr rig then buying accessories for an old digital camera seems like a really bad idea.<br>

All I need is raw pixels and a tethered view on my macbook for easier focusing - I will be using all the stuff from my medium format system and just attaching the digital bit to it via one of those adapters they make in china.<br>

Also to clarify, I already own a V system with a set of lenses and accessories so I need something that works with that. 99% of solutions out there are for thousands of pounds which is silly in my opinion in this day and age just as scanners are silly. Everybody mentions printing these small prints but what do people do with small prints?<br>

If I'm making an exhibition I need large prints, not A3 size stuff. For small prints it's easiest and cheapest still for me to just use an enlarger. All of this digital go around is to print big and to resolve grain in the digital domain. Grain is integral. If you zoom in and you see pixels instead of grain then to me that is not a scan worth of a print.<br>

I want to extend my current gear list into the large format domain rather than smaller formats than what I actually have. I don't do commercial photography so I don't really care about all these "pro" features as I can't pay the bills using them thus I am reticent about spending all this money on a current dslr. 1.5k quid is a lot of money guys for what is basically a single body and a single lens both used. I can grab two lenses for my hasselblad and at least three additional backs for the same money.<br>

Digital is not cheaper than film when you look closely, it is way more expensive and all this gear is loosing in value like crazy it seems. </p>

<p>It sounds like you already have a digiatl Nikon. I don't understand why you would spend hundreds of dollars rigging up Hasselblad stuff that's bulky, not made for the job, and won't do the work well just because it's bigger and more expensive, when one of the best micro/macro lenses in history, the Nikon 55/3.5, can be had for around $75, and the tube you need for another $50 or so.</p>

" I just want to figure out if there are any issues with just using the 35mm dslr body with my regular medium format

system."

 

No. Except that medium format lenses generally aren't designed for the resolution of small format cameras.

<p>Hi Borys,</p>

<p>I think I understand, you want to be able to scan your MF negatives using as much of your Hasselblad kit as possible and want to minimize your investment in the digital portion (the DSLR).</p>

<p>First a question - are you scanning all types of film - color negative, B&W and slides?</p>

<p>I am using my Canon 5D2 and 100 macro lens. It requires a minimum of 6 RAW shots and I end up with images of about 8500x8500 pixels or about 72mp (after cropping out the film border). That is about 25MB as a high quality Jpeg or about 195MB as a tiff file with no layers and no compression. One nice thing about using the Canon is that I can shoot tethered and can us the Eos utility to micro step the focus while watching a 10x live view image on the PC for both focus and framing.</p>

<p>Since I have started with slide film I have been very impressed with the colors, saturation, etc. I feel that the DR of the slide isn't that much beyond the old Canon (I know it is but just doesn't seem to be a big deal).</p>

<p>What I do find lacking is the sharpness. When I look under a loupe I think I am seeing more resolution that my DSLR scan. So I will be trying a extension tube on the macro lens to see what more I can get from it. I think I have ruled out camera shake, etc.</p>

<p>If I can obtain the sharpness I am after, then the next issue will of course be dynamic range for color negative and B&W film. That I believe is a place where just any old crop DSLR may not be ideal. This is where the Nikon D800's, D810's come in. But it is also where the Sony NEX 6, NEX 7 and the A7's come in. Those Nikons and Sony can capture more DR. Just one problem - all of those cameras listed above will set you back $$$$ except the NEX models.</p>

<p>So if you are okay with APS-C / crop take a closer look at the Sony NEX-6 or NEX-7. I currently have a NEX-6. Its a camera that has a fantastic image, just kinda crappy menus and speed. For slide scanning none of these issues should be a big deal (especially if they can be tethered). The good part, with Sony's throw anything and everything out there are fast as you can marketing approach, the NEX's are now going for around US$400 used these days! Then get a manual adapter to get your Zeiss to E-Mount (or use Zeiss to Canon and Canon to E-Mount).</p>

<p>Here I too would recommend following the sage advise of the others regarding the lens. Get one of the cheap, great macro lenses mentioned above. Maybe with one extension tube. Mount the NEX on your copy stand and shoot sony raw files. Perhaps you could get the NEX 6 or NEX 7, a great old manual focus macro lens and one ext tube for about US$600.</p>

<p>I think the bigger issues with going after more resolution is that when you have stitch so many images (say anything 6+) on photos that have large featureless areas (eg. blue sky) the stitching becomes a pain in the rear. To date, my nemesis slide is just a simple flag against the blue sky shot. There is so much nearly featureless blue sky that the stitcher never produces an image I like. I find that this is the only distraction on my mind when shooting film. This is where the D800, D810 and Sony A7r would come in handy since they are full frame high resolution so less overall images to stitch. I think I once calculated that using one those three cameras would yield something like 12500x12500 images or about 150mp!!!</p>

<p> </p>

<p>Reality check. No MF negative or slide is going to contain 120 MegaPixels worth of image data. 5" x 4" film maybe, but not even a 6x9cm frame is going to do justice to 120 Mp resolution. And even if there was that level of information on the film you'd need a copying lens with a resolving power of over 90 lppmm across the entire frame to capture it. That ain't going to happen on any sensible budget with a standard bellows setup. Maybe if you had a few grand to spare for a super resolution flat-field apo repro lens, but with a standard taking lens - no way.</p>

<p>120Mp across a 6x7cm frame represents a level of detail well under 6 microns, which is smaller than the size of the average dye cluster in a colour slide, and certainly below the level of subject detail captured in even the most carefully taken picture. Sure, it represents "only" about 4500 pixels per inch, and some drum scanners claim to exceed that. Well, they may output 4000 or more pixels per linear inch, but the evidence of your eyes should tell you that most of those pixels are empty data, since the grain structure or dye clouds of the film aren't even visible! Just the depth-of-field necessary to cope with the thickness of a film emulsion mitigates against that level of resolution. Stick a piece of film under any halfway decent microscope and try to focus on the entire emulsion depth at once - not possible.</p>

<p>Somewhere in the region of 3000 ppi is a much more realistic target. That's a scan size of about 45 to 50 Mp.</p>

<p>Thank you for that info Rodeo Joe!</p>

<p>I know Ken gets beat up pretty bad from time to time however I had read this article on his site (<a href="http://www.kenrockwell.com/tech/film-resolution.htm">http://www.kenrockwell.com/tech/film-resolution.htm</a>). The claim is that Velvia 50 will require about 320 pixels per millimeter. For 6x6, that would greatly exceed the 45..50Mp you stated.</p>

<p>That being said I just tried an extension tube (for 1.3:1) on my 100L and I don't think the image looked really any better than it did at 1:1. So I definitely believe I will not be getting any more detail from my 6x6 slides unless there is something else I am missing.</p>

<p>I am a bit bummed. I had hoped there was more detail in there and I had hoped I would be able to get it with a macro lens, tubes and a fairly modern DSLR.</p>

<p>Now I am wondering something, other than dynamic range - is a drum scan going to be any sharper than a DSLR scan with a macro lens?</p>

 

<p><strong>Dear Brad</strong>, you are definitely on the same page as I am and yeah dynamic range is something that does worry me albeit it is possible in theory to do a HDR type of technique at the reproduction stage, it should at least in my minds eye look fine when tweaked manually. There is really just one picture (if done correctly) for the HDR software so it should come out looking a lot better than HDR from actuall multiple photographs where the micro movements and differences seem to be the sources of “that obvious HDR look” - I’ve never tried all that fancy stuff in software so maybe you can’t avoid it in the end, I can’t tell right now. I will experiment with all kinds of technique once I actually get down to it nitty gritty and will find out.<br>

A possible solution for stitching troublesome shots is to simply stitch things by hand - you only need some sort of general marker for instance a tiny black paper marker you put on the film plane and then it ends up being in top right corner, top left corner, lower right, lower left of the respective shots you then put together based on that marker and then you also have a 5th shot without the market in the middle to replace the black spot. So if say you have 12 shots to composite together in the software, you could give the software 6 shots + 4(+middle) shots already composited into one where the big ambiguous area is located or just do all of it manually. In the end you have to judge by eye if the shots line up after its done anyway to some extent. The marker is just to help you get in the correct ballpark too, so it does not have to be pitch perfectly in the same spot on every shot. It is easier than it sounds - I’ve done similar stuff and it’s just a matter of patience and zooming in photoshop until you see nice big pixels. If the pixels colour data does not match 99% then it’s off, you move it, you check in a few places, you move again, check by eye how it looks in general and you finally get there after rinsing and repeating until it’s done.<br>

I know close to nothing about digital slr’s so I will definitely check out the sony line as I’ve read most of those sensors are made by sony anyway so the cheapest camera with the correct sensor should be the way to go for this particular purpose. I am also keeping an eye open for the MK II canon but it seems nikons in general offer a bit more value for money. I was eying the Nikon D7000 too, what do you about those for this particular purpose?</p>

<p><strong>Dear Rodeo,</strong> I respectfully disagree, I think that it is mostly marketing and I've been hearing that same spiel ever since ~9 MP was a big deal. Sure it ads up when we take the entry criteria chosen by the salespeople but the stitching (compositing) make it unnecessary to get specialty lenses or movements like some suggested - you don't need a technical camera, there is no perspective here to correct and the higher the magnification the less perspective there is really to worry about as the depth of field is down to millimetres anyway. If I can reach very high magnification levels and shoot a flies eye with my gear then why would I need a special lens to shoot a flies eye on a piece of film, the lens will see it almost the same if you set it up correctly - in theory, of course which is why I was asking about peoples experiences. This is even more so with medium format than large format as with a magnification of 2:1 which on large format would be a big pain in the ass, there is a lot less to correct optically as far as my limited knowledge goes at least.<br /> The marketing holds up when we take the requirements of yesteryear and the tech of today in a selective manner but if you are going digital there is no need to avoid excessive post production as that is inevitable at the DTP stage anyway. I can stitch and colour correct by hand on a 90ties era CRT and a low resolution mouse so I don't rely on a fully automated process with the much better tools I have today - the whole point of doing it all yourself makes sense really if you are in fact good in manual post because sure if you rely on everything being automatic it won’t work at all and you won’t get much information at all in there.<br /> Lens is not sharp anywhere outside the middle? You crop to where the sweet spot is and that is where you stitch. Colour reproduction is wonky? You colour grade it or whatever the english term for it would be - do it by eye if need be, do it separately for different parts of the shot. Difficulty getting proper focus? You focus in tethered view on a 27 inch screen and etc. etc.</p>

<p>It's not in the end about raw information for me but about reproducing the original so that each of the grains has a few pixels describing it and not having to up-sample for processing. All kind of downsampling or upsampling is imperfect and it will always be more comfortable for me to never have to do it rather than doing it with even the best gear out there.<br>

120 MP is not too much, that is what you need to get 300 DPI when printing 1.2mx1.2m if memory serves me right - my philosophy is to never upsample, never downsample so I want to reproduce already at my output resolution. That is not the commercial approach but I am not a pro - my main arena is music but I did study photography in school and I did enough of it to know that there is a ton of things one can do technically and in post that has not been done yet to death or known to work or not work. Running around with a small frame digital and shooting stuff for private clients makes me want to hang my self - some of my friends do it right now but I got so burned out of photography after I got my diploma that it took me 7 years to even think about touching a camera again so screw that and the canon baseball cap to go with it, bleh. I have never seen a medium size digital print that impressed me as a print itself more than a mediocre enlargement from a typical darkroom - it’s like looking at a vinyl compared to a cd. Cd’s are dead ugly to me and I am trying to satisfy myself as the creator first if I am to try and push my stuff onto others next!<br>

I want to excite people and small digital prints do not excite me - all of the painters I know paint enormous formats most of the time compared to what the photo world is doing so I am approaching this whole thing as a painter, painting with light. I am using modern tricks to avoid shooting with properly large <em>large format</em> which I do want to move into when I can afford someone to lug it for me :)<br>

When all I had access to was a 35mm I stitched many photos by hand to make it look as if it was done on a medium format system so now I actually own one so why can’t I make things bigger than life with that too? While people are saying medium format is not large enough for this or that, they are using 35mm’s themselves and claiming that the results can beat film up to X sizes which is exactly the same mindset of getting big out of small I try and apply. It’s ok for digital people to fake shallow depth of field because hey it’s in fashion so why can’t I do optical tricks to make you see 120 MP worth of picture on there? Just a while ago 10 MP was too much for some people - I won’t be satisfied until digital backs can do 160 MP at which point I will say “ok that is enough for me” - most companies seem to share that with me as they keep releasing denser sensors by the year.<br>

<strong>In the end I also have a pinhole and that is good enough for some of my needs - we reach for more because we choose to, not because it is necessary.</strong></p>

<p>I have been doing some more reading after Joe's comments. The first thing I have learned is that my trusty old 5D2 is probably not the weapon of choice for DSLR scanning of MF negatives. Its resolution comes in at 78.6lp/mm vs resolutions of 102lp/mm for the 36Mp Sony imagers and 128lp/mm for the Sony 24Mp imagers.</p>

<p>I took a peak at the photozone lens review for my Canon 100L macro. They show the charts for MTF50 in lp/ph.</p>

<p>How does a person determine the lp/mm number for this lens (or any lens)?</p>

<p>I am curious, does it resolve well enough to hit the 45-50Mp number that Joe indicated would be more realistic?</p>

<p>And I think I now get the part about the lens needing to producing those resolution figures across the entire frame. I guess I probably knew it was somewhat of an issue but hadn't got far enough along yet to see it's effects. For my case using the least number of stitched images meant there would be more regions where the sharper part of the image from one frame would meet the lower resolution portion of the image from another frame. Now I wonder if that may have even been noticeable printed at 13x19?</p>

<p>So to mitigate this if one only has standard lenses, then many slightly overlapped images would be required and then force the stitcher to use the sweetspot (ie center) of each (good luck there). Or shoot all the images, crop to the centers first and then stitch.</p>

<p>Which brings us back to Borys proposal. He proposes to use his Zeiss MF lens on the APS-C DSLR. In doing so he is effectively pre-cropping (quite significantly for that matter) on the APS-C imager to use the highest and most consistent resolution portion of the len's image circle.</p>

<p>Given he intends to use a high enough resolution APS-C imager: is the Zeiss MF lens good enough (in its center when stopped down) for a final stitched 50Mp image? </p>

<p>If not, is there another cost effective lens option (which through cropping) might give acceptable even resolution across the APS-C frame. Perhaps an older Medium Format macro lens?</p>

<p> </p>

<p>The lenses I have Brad are top of the line at 3000 dollars per new lens and that coupled with exactly the fact that most lenses are sharpest in the middle I just do not understand the logic behind thinking that a small frame medium price lens will top a 180 CF Zeiss. Sure a 3 grand worth of canon lens will be great but I can't see how a 800 dollar lens produced today in china will be somewhat better just because it is intended for digital. It is intended for digital and a legion of people who can't tell the difference too so there are statistical and market elements going against the possibility that these new sensibly priced lenses are better than the old top of the line ones.<br>

The fact that there is cropping is only helping the MF lens as you have suggested. The detractors are just repeating marketing spiel - I mean no one beside you even understood the original question fully so I think you should not worry either about this silly 50MP limitation as that is purely conjecture as is the idea that digital lenses are better than film era lenses on a general basis. Whoever came with that should contact all those movie people who use all those old film lenses on their digital cameras. <br>

50 MP is only 60x60 cm at 300 DPI or roughly 24”x24” and at these sizes it makes more sense to just buy a digital MF back that does 50MP natively or to use an enlarger and print in the darkroom if sticking with film. It is easier to print in the darkroom than to produce an excellent digital print, DTP is really difficult and really expensive. You need to do a lot of proofs if you are renting a facility before you can actually do the print and you need top of the line facility to get the same dynamic range as you get with a darkroom print and that same concise quality rather than ugly inkjet print. Quality printing paper costs a lot of money too. It is cheaper materials wise and faster to print in the darkroom, even on the most expensive photographic paper. The limitation of the darkroom is mainly print size and no digital post.<br>

The only reason the limitation of film is 50MP according to so many people is because they own digital cameras that are about that resolution and read that idea or where told such a thing when buying their gear. The moment phaseone releases a 160 MP back, suddenly the limitation of film will be allowed to go to 80MP by the marketing people so that “twice as much as on film” can be used to sell more outsourced, automated camera kits.<br>

People also own drum scanners worth tents of thousands and labs which did not get on with the times and offer below 30MP scans for hundreds of dollars so of course the status quo is that film has only as much resolution as whatever all these people offer to extract from it.<br>

Photographers are also in my experience lazy when it comes to processing and are comfortable sending off their typical work load to be processed and printed by the same old people they used for the last 15 years and fear having to actually get up the sofa and do stuff themselves because the new age has dawned.<br>

God forbid some guy gets 80MP-100MP of information for a few hundred dollars and has a bigger digital penis afterwards in the minds of those who paid 60k for the brand new digital backs.</p>

<p>All of those people sure about how they are getting 30 MP of information with a 30 MP small frame DSLR - have you considered that your photography does not warrant 30 MP at all and maybe in fact closer to 12-16? See this sword cuts both ways so how does that feel to be doubting your skills as a photographer? Having resolution to spare in digital is a must especially when doing post processing while in analog you can ride the event horizont of technical limitations at all stages without much issues. In the same sense in analog audio you can overdrive most devices while in digital you need to keep the levels controlled because any digital distortion is square wave clipping so sure digital audio also is amazing on paper but one third of the reported dynamic range of digital in practice has to be left alone in case there are spikes in the material unless you want to use brickwall limiters on your converters. Sure the end product can use most of that amazing dynamic range but that is the final product, not during the recording process in a typical creative situation in the studio. Similar thing applies to digital post production, different rules apply to the process than to the finished product.</p>

<p>Who is to say I am not getting 95% of what say an ektar 100 can offer or an ilford delta 100 in terms of resolution - that is purely in the hands of the photographer. If that would be even just 70MP - 80MP then getting 120MP is about having the necessary digital headroom and workflow so that each individual grain is resolved in a bunch of pixels rather than a single one which does not translate at all. A single grain or line of grains does not equate a single pixel or a single line of pixels. 720 worth of lines of grains projected from an analog projector would look stunning compared to 720 worth of lines of pixels projected from a digital projector. That is because it is harder to trick the eye with pixels for whatever reason, they are clearly defined while analog media is reliant on many variables also in the viewing stage. So to recreate that into a digital print I want to print film grain and let the film grain compose the picture with similar variable taking place - the eye sees the grain but the brain sees a picture.<br>

That is the issue with pixels, a single grain is not limited to 255 values per colour channel and is not square but has an actual 3 dimensional structure so extra resolution is in my opinion anything but wasted. People are making 8k cameras and tv’s in japan while I am still watching 360p video on the web and only started thinking about moving to shooting 4k video in a many years once it is a standard. The trends are against the idea that we have enough resolution.<br>

It is true that this grain structure is not always chemically formed to be all that we expect it to be but resolution is about covering the possible requirements of the material and not specific requirements for each shot or technique - there is just no need to limit resolution like that for post processing given todays computer hardware. If I can get 120MP and be covered from head to toe that I am getting enough information out, then why would I settle for 50MP and then upsample it to 120MP for printing?</p>

<p>Boris, I support your effort for more pixels. I've done large prints from multi-exposure stitches.</p>

<p>For a high-resolution capture from film:</p>

<p>- Use a dedicated macro lens. For your project, to get 120 MPx, you'll be making several shots. Perhaps 6 shots at 1:1 on a DX sensor DSLR. For the highest quality, you'll want a lens optimized for 1:1; this would be the Olympus 80mm bellows lens or the 75mm APO Rodagon D 1x. The usual macro lenses are optimal at more like 1:3 or 1:4 and normal lenses are way out of their optimum range.</p>

<p>- Or, perhaps you'll make more shots at 2:1 on a DX or FX camera. For this, a macro lens reversed will be pretty close to optimal. I'd be inclined to use a manual focus lens, the 55m f/3.5 or f/2.8 Micro Nikkor.</p>

<p>- You'll have to find a way to illuminate the original consistently, and shift to capture different parts of your original. Also, you'll need to hold the original flat.</p>

<p> </p>

<p>Boris, replying to your last comment:</p>

 

<blockquote>

<p>The lenses I have Brad are top of the line at 3000 dollars per new lens and that coupled with exactly the fact that most lenses are sharpest in the middle I just do not understand the logic behind thinking that a small frame medium price lens will top a 180 CF Zeiss. Sure a 3 grand worth of canon lens will be great but I can't see how a 800 dollar lens produced today in china will be somewhat better just because it is intended for digital.<br>

</p>

</blockquote>

<p>A $100 55mm f/3.5 micro Nikkor will be better for your "scanning" task because it is optimized for a subject-to-image ratio in the range you are intending. Your fabulous Hasselblad glass, and it truly is fab, is not optimized for macro focusing. I suggest it's worth the $100 gamble to try this. </p>

<p>By the time you resolve this, your images could have been drummed scan and you would be done. My time is worth money, and you say you want BIG prints. Go with the drum scan. The rest is making something simple much more complex and time-consuming.</p>

<blockquote>

<p>then the next issue will of course be dynamic range for color negative and B&W film.</p>

</blockquote>

<p>Brad, color negatives consist of a small dynamic range, so that's not a problem. Everyone talks about the huge dynamic range of colour negative film, but don't let that confuse you. There's DR in, and DR out; the film <em>captures </em>a large dynamic range in the original scene; but it <em>stores</em> it within a density range which is quite modest, and well within the range of a good DSLR or CSC sensor if you are using it to "scan" the negatives.</p>

<p>B&W negatives are a different matter, and may require HDR-style shooting to capture the full range. </p>

<p>"The claim is that Velvia 50 will require about 320 pixels per millimetre." - ROFL!<br /> When I was working producing photomasks for the electronics industry with a 2 micron feature size <em>true </em>resolution limit, we needn't have bothered with special Lippmann plates then? We could just have snapped away with a cheap Leica loaded with Velvia. Yeah, right!</p>

<p>"B&W negatives are a different matter, and may require HDR-style shooting to capture the full range." - Only if they're overdeveloped. The Dmax of a properly exposed and processed B&W negative should rarely exceed 2.2D above base+fog (~ 160:1), which is well within the dynamic range of any modestly-priced DSLR. In fact it's expanding that low contrast range into a viewable higher one that loses you digital information and tends toward a posterised effect.</p>

<p>Edit: Re drum scans. I've yet to see any drum scan that shows the grain or dye cloud structure of film. The flying spot size can't be got down to the diameter needed for that kind of resolution. Hence my earlier comment about "empty pixels" - i.e. pixels containing little or zero different image information from their neighbours.</p>

<p>And Borys, when you get down to the sort of resolution and magnification required to image individual dye clouds or silver "grains" (not really grains) your depth-of-field isn't measured in millimetres, but microns, and is generally far shallower than the thickness of a film emulsion. Unless you stop down far enough for diffraction-limiting to take over.</p>

<p>Hey Guys,<br>

I've been hitting this subject hard since my last post and I've been making slow progress in maniacal leaps of 8h shifts of researching, building and testing.<br>

Some of the issues I need to resolve right now is a better light source, newton rings and oddly enough the biggest offender at the moment which is surprisingly PS photomerge. The blending is really bad and the alignment only gets within 90% so I get better results by hand atm as there is no easy way to fix the remaining 10% when done on auto. Minor things include support stability (vibrations) and dust which first require a better room and better tethered support for the dslr.<br>

Issues I've already resolved to some degree is mounting the camera on a low vibration wooden tripod and mounting the light source on a table I've made to purpose, aligning the film plane with the focal plane of the digital sensor within the mechanical limitations of the mounts responsible for both planes with measurement, managed to setup tethered shooting for the d3200 (quality histogram) on my mac (a lot of lost hair) and managed to create an already very good preset in cameraraw for processing colour negatives.<br>

I've bought a brand new nikon d3200 to work as my digital back since I was tired of waiting for a used D600 and I am determined to make the rig work first with that and then once the physical aspects are sorted I will choose the optimal digital element for the setup. I'm using a shitty hassel-nikon adapter and the rest is my hasselblad setup (which is good stuff imho). I am keen to buy an old digital back for the hasselblad rather than another small dslr - once the d3200 is no longer nessesary I will grab a cheap matte rig and turn it into a video camera.<br>

For any one interested you can definitely make it work with a d3200 which is dirt cheap compared to say a Nikon d600/d800 or the comparable canon d5 mk II/mk III. Tethered shooting can be done but not with live view, you can hack your way into live view but it is at the lcd resolution of the camera itself so I'd rather do my initial focusing with the eyepiece.<br /> In practice tethered is downloading the RAW's in a few seconds so I just take say 20 shots and judge focus that way when setting up and then might correct focus for specific shots midway if I see it gets too far out of whack.</p>

<p>I take 9 shots at 24MP at the moment purely because I use my bellows accessory which starts at a pretty large magnification given the adapter + crop already present. I think the auto stitching would be better at 6 shots but for that I need some extension rings which I am trying to buy right now used. At the current magnification levels that requires the 9 shots minimum I get about 160 MP which is workable therefore I will try and make the rig work at those magnification levels and then at a 2shot stitch magnification level for general archival/selection tasks. The auto stitching should do fine with just 2 shots I hope.</p>

<p>According to my current experience the most important element is film flatness and newton rings - I have yet to get it flat enough so that the limitation is the lens. Light source heats up the film so I doubt anti newton glass would help unless the film is sandwiched in it which means another optical element besides air between the film and the lens. I am worried about trying to clean the anti newton side of the glass - any ideas wether this is possible to do with just a microfibre and zeiss cleaning fluid?<br>

As for lens <strong>I am eager to use the 135mm zeiss</strong> for the hasselblad as it is dedicated to be used with the bellows for high magnification levels. Currently I'm using the good old 80mm one with the bellows and waiting for a good deal on the 135. <strong>I appreciate the recommendation for the specialised macro lenses - I suspect the 135mm would be better though since I do the shooting at higher magnifications than 1:1 - am I wrong here? I am relatively new to these kind of magnifications.</strong><br>

<strong> </strong><br>

<strong>In regards to "you could send it off to drum scan"</strong> - the problem with drum scan is that very few places actually own top of the line drum scanners and have the staff to operate them. The best I can get in my area is a coolscan 9000 which is not a drum scanner at all and in such a case I’d rather buy the scanner for my self. If someone would scan my negatives with fluid mounting on a hundred thousand dollar worth of a drum scanner and do it expertly without destroying the negatives at less than 50 dollars a roll including postage then it would make sense but otherwise I am not convinced.<br>

I don't see value in scanning a whole roll at high quality - I might have 5 rolls and want to extract just 1-2 frames from each of those 5 rolls due to my practice of shooting "safeties", some on different film, some bracketed positives etc. etc.<br />I've spent 500 dollars till now including the nikon d3200, if I manage to get the shots sharp enough below 1000 dollars total cost I am beating the nikon coolscan 9000 by thousands of dollars for price of admission alone.<br />For the extra 500 dollars I am sure I can at least fluid mount and improve the light source - I'm buying a kaiser one soon for about 80 dollars. I've made a lens shade out of ready made hydraulic tube with a rubber seal which had the ideal dimensions for the 80mm zeiss, it cost me about 2 dollars!</p>