Which macro lens to get in order to scan 120 film with a Nikon D600

<p>Hey Guys,<br>

A while ago, I found this excellent article and I have high hopes of using it for myself:<br>

http://petapixel.com/2012/12/24/how-to-scan-your-film-using-a-digital-camera-and-macro-lens/</p>

<p>As I am not much of a digital guy as far as cameras go and I don't know much about what is available for the nikon mount - Can anybody recommend an economically sound lens choice to be used with a Nikon D600 in order to scan 120 film?<br>

The D600 belongs to a friend of mine so I'd rather not shell out for a specialised macro lens that everybody covets just to scan film if I can make do with a cheaper one as outside of scanning film it will sit in a cupboard.<br>

I will be scanning square format so ideally I would like to get a lens/lenshood combination that has the correct reproduction ratio and distance from film combination to divide a square 120 cell into 9 shots with the D600.</p>

<p>Alternatively I would be happy to buy a more expensive Hasselblad V mount lens to use with an adapter with the Nikon although I have no idea off the top of my head if the "focus included" 120 macro for the hasselblad can get sharp at the distance afforded by a lens hood as I do not own one at the moment. I do have a bellows so I could get the more macro of the macro offering for the Hasselblad but making a tower out of the bellows extended seems like a PITA. Also those lenses seem to be a bit overpriced when sold without the bellows at the moment.</p>

<p>Big cheers for any ideas!</p>

<p>Crumpets,<br>

BP</p>

<p> </p>

A 55mm f2.8 or f/3.5 Micro-Nikkor, or a 105mm f/4 Micro-Nikkor will be fine for your use. Manual focus of course, so use

the camera's live view fubnction to focus with.

<p>Should we call this scanning?<br>

The old AI micro Nikkor lenses would be good and are much cheaper than the AF version. You don't need AF for this application. <br>

I personally use the Nikon PB-6 bellow with a number of enlarging lenses. </p>

<p>The maths to discover the needed reproduction ratio (RR) is quite easy. The average size of a so-called 2 and a quarter inch square frame is 56mm x 56mm. So if you want to make 1/3rd of that frame height fill the 24mm side of a D600, then you need an RR of 24/(56/3) or 72/56:1. Windows calculator makes that ~ 1.286:1. In other words, just over lifesize. There's no Macro lens readily available from Nikon or any 3rd party lensmaker that will go above 1:1; meaning that to get what you want you'll have to add a bellows or extension tube.</p>

<p>However, if you lower the requirement just a little and take 1:1 segments of the 120 frame, then there's a whole raft of lenses that will do the job without need of a tube or bellows. I can personally recommend Tamron's excellent 90mm f/2.8 SP AF macro lens. It focuses to 1:1 straight out of the box. Or the excellent 105mm Micro-Nikkor will do the job just as well, but costs a lot more.</p>

<p>If you're on a shoestring budget then a manual focus 55mm f/2.8 AI-S Micro-Nikkor can be got for very little money, but you'd have to add the cost of an M, M2 or PK-13 extension tube, since this lens only focuses to 1:2 unaided. Any of the aforementioned lenses will have more than sufficient quality to do the job - provided you provide them with a rock steady support and a perfectly flat frame of film to copy.</p>

<p>Having said all that, IMHO you're pretty much wasting your time taking segments of a 120 film frame. The gain in detail will be absolutely minimal compared to simply copying the whole frame onto digital in one go. At most, I'd say that filling the D600 frame side to side (in other words copying at a ratio of 36mm to 56mm) and then stitching two frames together should be more than good enough. Film doesn't hold nearly as much information as some film fanboys would have you believe.</p>

<p>PS. The poor quality of the V700 shots shown in the OP's link is down to the crummy bit of bottle-bottom that Epson use as a lens. Nothing to do with the number of pixels captured.</p>

<p>I'm doing this job with a 55/3.5 Micro-Nikkor. I have two versions of that lens, and the modern 60 AF-D, and in this job they are all equal at their optimum aperture, which is about f/8. All varieties of the Micro-Nikkors are some of the best lenses Nikon has ever made. You can pick one up on Ebay for $50-$75, so why not? Beware of the f/2.8 version--these are prone to oil leakage onto the aperture blades. Before switching, I tried this with several different first-rate enlarging lenses, and for this particular job, working at close to 1:1, the Nikkor blows them out of the water, both in contrast and sharpness.</p>

<p>For 6x6 doing one shot (non-stitched) you won't need a tube; for 35mm, you will. I find that 6x6 is about the cross-over point to a flatbed--an optimally set up flatbed will get close to camera scanning, but not quite there. A two-shot camera stitch will be much better than any flatbed. Microsoft ICE is a great stitching software for this. You may find a single shot, non-stitched, to be totally adequate, too.</p>

<p>Here's my setup: Almost all of the B&W on my Flickr pages was done from negs with this rig. You can take a look at the original size versions there.</p>

<p>Many of the flatbed film scanners do an good job with larger format images, cost relatively little, and are a LOT faster than the multiple image method in the OP's article.<br>

I find Michael's summing up here to be rather rosy, and I have done it both ways.</p>

<p>JDM, I'm curious what you find rosy. Not being combative, but looking for improvements on my flatbedding :-)</p>

<p>The problem I've consistently had, through several generations of flatbeds, was smeary sharpness that I was able to better with a camera and the right lens every time. With enlarging lenses, the flatbed might beat the camera. . . it's that close. I'm happily wet scanning 4x5 on a flatbed, but with a D600, a definite bump up from my D300, the choice would be obvious for me for 35mm, at least, and probably for 6x6.</p>

<p>For one-shot, not stitching, the camera method is extremely fast, too--more so that a scanner, for sure.</p>

<p>This is the best I've been able to get with 35mm/flatbed:https://www.flickr.com/photos/mdarnton/11182722354/sizes/o/<br>

vs the best with a camera (focus is on the statue, not the person):

<p>If we are talking about black and white 120 film or positive 120 film I agree with most of the comments and I add support to the 55mm f3.5 suggestion. Yes it is a great copy lens and it is inexpensive. Note Michael's warning about the f2.8.</p>

<p>However if we are talking about colour negatives I disagree with this approach. Dedicated negative scanners use a different concentration of RGB then cameras do. Forgive me if my explanation is not very scientific or even accurate, but the upshot is that it is damn near impossible to get your colours sorted out properly from a photograph of colour negative film. If indeed we are talking about colour negative 120 film rent one of those very nice Nikon film scanners for the weekend and scan it all that way.</p>

<p>I suspect a lot depends on the flatbed film scanner one has, but I've never found even two stitched together images to be as good as my flatbed scan on a Canoscan 9950F or a 9000F.<br /> It's also a question of efficiency. For a few slides, camera copying can be good; but if you're doing 5,000 images or so, it's another matter.</p>

<p> </p>

<p>I think the "oily blades" reputation of the 55mm f/2.8 has been greatly exaggerated. Like any fault, it'll get reported, while non-faults (obviously) don't get reported. The result is a distorted view of the reliability of the item in question, and I think the 55mm f/2.8 Micro-Nikkor has had a particularly bad rap in this respect. Besides, it's an easy thing to test the lens by flicking the stop-down lever and seeing if the iris shows any tendency to sluggishness or lag.</p>

<blockquote>

<p>"Dedicated negative scanners use a different concentration of RGB then cameras do."</p>

</blockquote>

<p>I don't think so Phil. The light source stays the same colour in most scanners, and indeed it's not possible to change it if it's the common fluorescent tube type. All that the scanner software/firmware does is change the gain in the RGB channels, and this is exactly the equivalent of altering the white-balance of a digital camera. The only major difference is that good scanners use a 16 bit A/D conversion, while DSLRs are limited to 12 or 14 bits. This can make a slight difference to the highlight quality when scanning negatives, but in most cases the low density range of negatives doesn't present a problem to a 12 or 14 bit converter.</p>

<p>If there is an issue, the mask colour can be quite readily overcome by adding a complementary blue filter to the lens of a DSLR and increasing the exposure to compensate. Agreed that the post-processing to achieve a decent colour balance is more difficult with negatives, but I've yet to see a dedicated film scanner that can do a perfect job straight off and without some manual adjustment.</p>

 

<p>I think capturing 120 film in a single image on a 24MP camera would be preferable to capturing the image in multiple shots. For critical work, a macro lens will be better than any other FX camera capture option. Since AF isn’t needed, a used 55mm/2.8 or 105mm/2.8 manual focus Micro-Nikkor will work well. Alternatively you could rent a lens from lensrentals.com or borrowlenses.com if you are in the US. There probably are similar rental companies in Europe too. The Nikon mount Zeiss 120mm/2 Makro Planar ZF.2 would be a good rental choice. These lenses all focus down to 1:2 reproduction ratio, which should be adequate for 120 film capture.</p>

<p>JDM, regarding speed, it's worth noting that this method was developed by libraries and archives (which is where I first started reading about it), who were finding flatbeds too slow.</p>

<p>I think, not sure, that the problem with making color conversions isn't the light; it's the software and how it deals with the orange mask. I don't scan color, but people who were using cameras finally arrived at ColorPerfect as the best way to convert color camera scans. I don't know if ColorPerfect is still available--I think they rolled it into one of their other products. As I understand it, the real problem with the mask is that it's not linear--it's also made photographically in proportion to the underlying subject matter, thus the problem of filtering it out, directly. That's my understanding, anyway, perhaps faulty.</p>

<p>If you want to see what camera scanning can do, go to http://shorpy.com They are putting up stuff from and done by the Library of Congress collection, which is all high resolution copies of original negs done with 4x5 cameras and digital backs. Lots of WPA stuff, but tons of other great photography, too!</p>

<p>Joe, I will take your word for it. I can't find the article that explained to me why colour correction of negatives is such a pain. I am very comfortable with Photoshop yet I have not had much luck with this but I do see that others have done a better job than I. I would just warn the OP that this is an issue. Here is an article where they come up with some very nice results, but look at the process, it's more complicated then setting WB.<br>

http://petapixel.com/2012/05/18/how-to-scan-film-negatives-with-a-dslr/</p>

<p>I recently purchased a AF-S Micro Nikkor 60mm 2.8 G and have found it to be the best micro Nikkor lens I have ever owned. It is clear, razor sharp and renders beautiful color.It is sharp corner to corner even wide open. I an sure copies vary in quality but this is one I would at least give a try. I admit it is a bit pricey, but worth it. It has also served me well as an excellent portrait lens.</p>

<p>-Owen</p>

<p>You will not need a 1:1 macro as the reproduction required will be less than the 1:1 required to reproduce a 35mm format slide of negative. Any good macro lens will do the trick, although for best reproduction a high quality repro lens is best, such as the Rodenstock Apo-Rodagon series.</p>

<p>I use a Canoscan 9000f https://www.imaging-resource.com/SCAN/CS9000/9000F.HTM and am completely happy with it - though I should add that I use it in conjunction with Photoshop which allows me to bump up contrast, vary colours etc.</p>

<p>There is another thing I'd consider; even if with a DSLR it may work faster frame-to-frame (and indeed flatbeds aren't fast), if you have to digitize a lot of negatives and the D600 isn't yours, it's awkward. If you want to do a lot of negatives, you'll just have to hope your friend can make the D600 available frequently (if it's just a few, much less of an issue, obviously). A flatbed you get yourself - which can be had for relatively little money - is available whenever it suits you.</p>

<p>That is a lot of excellent information over here guys so a big thank you to you all!<br>

Fortunately the friend in question works relatively close by so I would probably just have him use the lens as part of his setup day to day and keep my scanning rig at his studio rather than mine and just come over with my laptop and reproduce away. If the workflow does indeed flow I would grab my own body after I buy things higher up on the list.<br>

I am, unfortunately working on a "shoestring budget" but in the sense that my "studio" is more of a music studio with some design capability that I am equipping now for photography as I am no longer affiliated in anyway with any institution that used to offer me such facilities in the past. It all ads up very quickly and none of the bits and pieces I had left from when I studied photography seem to satisfy my current thirst for quality. A flatbed might be convenient but is a very specialised piece of gear and cannot offer me enough resolution for what I want to do at the prices asked - I'd rather shell out that kind of money for an older phase one digital back.<br>

I want to stitch many shots together, partially because as I will be printing large sizes for exhibiting and I only have contact with relatively small galleries thus people will be nose up close to it at times - but mainly because I want to resolve the grain with a lot of resolution to spare and grain will be small too given spot metering, analog filtering and slow speed film. So I wish see the limit of the film on the print rather than of the digital reproduction.<br>

While larger prints should be viewed from further apart I find it artistically important to make it hold out close by - especially when I am trying to make it work the same way a painting does without making it obvious that there was post production (a layman nowadays can tell a print is pixelated when up close). So ideally I would like to be able to go with ~72x72” at 200 DPI which is 14400x14400 pixels (~207MP) so with careful technique, 9 shots at 24MP stitched together should give me the raw resolution necessary. Now I am not worried about having too much resolution as again the grain is an integral part of what I plan to do with these prints and the post processing will be rendered natively at this resolution.</p>

<p>I will try and see what I can buy out of the lenses recommended and will probably need to grab some extension tubes to get the necessary reproduction - any particular recommendation regarding this? I am not worried too much about colour correction as my digifoo used to be up to scratch when it comes to post processing last time it mattered and I will have to spend a lot of time doing DTP anyway as I will probably rent a day at a facility rather then trust the printing to someone else.<br>

Again guys, a lot of excellent information in this thread so I have high hopes for making this work for me too!<br>

I will make sure to share some results - might take me a month or more still since I am servicing my gear at the moment and still waiting for orders and buying bits and pieces. </p>

<p>Crumpets and tea,<br>

BP</p>