How Good a Macro Lens do you Really Need for DSLR/Mirrorless Camera Scanning?

The question about what macro lens to buy for camera scanning is one that comes up time and time again – it’s certainly a question I get asked regularly. I have wanted to write an article about how to choose a lens for a while, but I’ve found it a difficult subject to tackle without getting too bogged down in the minutia. The issue is, the answer to this question is far from definitive. That said, there are some folks out there that have done some fairly extensive testing on a wide range of lenses to demonstrate specifically how good certain lenses are for the task of digitising film with a digital camera. Having discussed this topic with one such photographer, I have now found myself with the resources I need to lean into this topic and hopefully provide some useful insights for anyone looking for an answer to this question for themselves.

The photographer in question is a chap by the name of Richard Karash. Richard is well known in the online film digitisation communities as someone who has painstakingly tested a very wide range of macro lenses for the purpose of digitising 35mm film at 1:1 with a full-frame camera. He has then shared many of his results on various Facebook groups and forums around the internet. Much of this information is out there and relatively easy to find too. Here is a link to a thread on RFF that lists some of the best lenses Richard has found divided into tiers: ‘What is the best lens for 1x Camera-Scan?‘ If you’re going to just dive into that thread, then it’s worth noting here are slightly different considerations for medium format as well as for digitising 35mm with an APS camera. That said, this should still provide some useful insight for anyone getting into this subject for the first time.

Job done then, right? If those are the best lenses, then just pick one and get on with digitising your film. Well, actually, I think there is more to this problem than just attempting to seek out the “best” lens. For a start, your camera might not be compatible with any of “best” lenses, or at the very least they might be awkward to mount. It’s also the case that whilst Richard ignores the really expensive copy lenses, some of his listed lenses are still pretty expensive – especially, unsurprisingly, some of the “best” ones. And then there’s the individual expectation factor – one person’s “brilliantly sharp” might be the next person’s “less than adequate”. Finally, and this is much more key to the point I want to make in this article, the “best” lens might not even be required by a very large percentage of people who use digitise their film.

This is, in fact, something that I find a little frustrating about the world of photography these days. It seems people often fall into the trap of thinking they need the highest resolution and the most amount of pixels to get the absolute “best” results. This is something that can be found a lot in the world of digital photography, but it is rife in the world of analogue photography too, especially when it comes to digitising film. This has always seemed somewhat ironic to me, especially where an extremely large percentage of people who digitise their film do so for display online or printed fairly small within a ‘zine where high resolution isn’t required. The fact is, unless you’re a fine art photographer displaying your images printed to a very large size, it is quite unlikely you actually “need” an amazingly good lens and a very high pixel count camera to digitise your film.

With this in mind, I recently got in touch with Richard asking him if he would like to be part of an article for this website. He kindly agreed, and so in preparation for this article, I asked Richard a few questions around his thoughts on how good a lens is actually required for camera scanning. Now, this might sound like an odd question to ask someone who has compiled a list of lots of lenses for this purpose and organised them into tiers by quality. You might expect him to just refer me to his lists and suggest that buying one of the best lenses off the list would be the way forward. Actually, whilst Richard has done a lot of tests using test charts and very high-quality equipment to determine what is “best”, he also has some much more practical and pragmatic views on what is required to get a good result. He makes a point of highlighting this in the preamble to highlighting lenses in the article linked above.

In fact, me getting in touch with Richard spurred him on to write another forum thread titled, ‘How Great a Lens do we Need for Camera-Scanning?‘ in that thread, Richard compares one of his “Top Tier” i.e. best quality lenses to a lesser 3rd or 4th tier lens, i.e. a lens that under test conditions doesn’t perform particularly well. The key is though, as he demonstrates, just because the lens doesn’t test well, doesn’t mean that it can’t provide adequate results depending on the requirements of the person doing the digitisation.

With Richard’s permission, I am going to share his comparison as well as some of his thoughts, combined with some of mine, to hopefully demonstrate both the objective and practical difference between an excellent quality lens and a just about average quality lens when it comes to the purpose of film digitisation. Some of this can be found in the forum thread, some of it Richard has since shared with me via email for the purpose of this article.

Comparing Lens ‘S’ to Lens ‘V’

To make the comparison Richard digitises the same image, an image of the famous “Duomo” in Florence. On the transparency (film) is sharp from corner to corner. This image is the photo he digitised with the two lenses:

Richard doesn’t initially make it clear in the forum thread which lens he used to digitise the above version of this image saying that “At this resolution, 1000 pixels, we won’t distinguish the two lenses.” The point here is that if you’re only ever going to show your photos online, they are unlikely to be much bigger than the above image and therefore even an average lens is more than adequate for this purpose.

He then goes on to share two versions of the image in higher resolution. JPGs at 5500 pixels wide, unsharpened and about 10MB each. He calls the lenses ‘Lens S’ and ‘Lens V’. You can compare them here: Lens S 5500 pixel file vs. Lens V 5500 pixel file

I think this actually makes an interesting first test for any individual’s requirements. Before we get onto specific pixel-peeping, have a look at those two images in their larger forms and see how much difference you can tell between them. If you can’t see much difference then you might have answered this question for yourself already…?


Of course, the comparison doesn’t end there. If we really want to see the difference between these two lenses, we need to look much closer! To compare the lenses in more detail, Richard next resorts to pixel-peeping, i.e. zooming right in on the image. Here are some crops of the above image. The first image is from the centre of the frame. As you can see, there is very little between them:

This second image is a crop of the lower edge of the frame. In the lens V image, we can see the quality beginning to fall apart a little. It’s not terrible, but the grain is definitely not as well rendered and the detail in the image a little smeared.

In this final image of the lower right-hand corner of the frame, we can see that even the excellent ‘S’ lens isn’t as sharp as it is in the centre. The lesser ‘V’ lens is also a lot more smeary with quite a lot of loss of detail. One way or another, there is a lot of difference between the two lenses here.

At this stage in his forum thread, Richard reveals the two lenses:

– Lens S is the 70mm f/2.8 Sigma Macro ART, his go-to lens at around £450/$550.
– Lens V is the 75mm f/3.5 Vivitar “Flat Field Lens”, an enlarging lens that can be picked up off eBay for not much money.

So now we can see and have proof that the Vivitar lens isn’t as good as the Sigma. Even if it wasn’t as obvious in the 5500-pixel images, can definitely see it in the crops above. That is to say, when we pixel-peep, flaws become a lot more obvious. The question remains though, just how much worse is the Vivitar when really tested?

In the next images, we can see how these two lenses compare when assessed with a Vlad’s Test Target. In case you aren’t already aware, these targets are very helpful for judging image quality right across the frame as well as for checking alignment (all four corners equally in focus), and, of course, for comparing lenses. You can read more about them here.

In an email to me when sharing these images Richard said that with one of these he, “…can test resolution across the full-frame.” and that to conduct this test he “… mounted [the target] in ANR glass for flatness and carefully aligned.” Here is the comparison for the two lenses at the extreme corner, first at 100% where we see resolution difference and some chromatic aberrations (colours around edges of areas of contrast) in the Vivitar:

For a really serious comparison of lenses, here is a look at 400%! Just to repeat that number, we are now looking at 400% crops folks, this really is pixel-peeping! Again, this is an extreme corner:

Of course, at this stage, we have fallen off a pixel-peeping cliff. Looking at results at 400% tells us a lot about the quality of a lens. We can see these issues in much greater detail than we can when we compare the two images at 5500 pixels. We can also see the difference to a greater degree than when we looked at the crops. Looking at these test target crops, you could be forgiven for writing the Vivitar lens off as a load of crap…

But… the reality is, we would have never seen these issues when looking at the first 1000px image. And, crucially, as Richard goes on to say, “the inferior lens makes a file that will work fine for many uses, and will make an acceptable 12×18″ print.” In short, in Richard’s opinion, if you are printing your images at 12×18″ or less, or only displaying them online, then the inferior lens might just do the job for you!

In fact, I would go even further than that still and say that you could print much bigger from the lesser-quality image and still get an image that many would deem acceptable. The reality is, that the bigger the image is printed, the further you tend to stand away from it. I.e. if you printed an image from lesser lens digitisation, you would only see the issues if you stood really close to it and peered at the corners… and how often does anyone actually do that when admiring a photo…?

Richard goes on to confirm that for archival scans or for “perfect” prints, yes, the more expensive lens does a much better job. I would agree with this too. But, I firmly believe that you really need to ask yourself what you are going to be doing with your images before you decide that you need a lens of that quality! Are you making important archival scans or perfect large prints? Or are you just digitising your images for display online…?

Also, it’s worth noting that none of this so far has asked the question about how good the lens that took the photo in the first place was! How good are your lenses? Are they all perfectly sharp into the corners? The massive majority of lenses produce images that are softer in the corners! I think it is also worth thinking about whether or not you take photos of subjects that require perfect corner-to-corner sharpness? To create an image that would be useful for testing, Richard needed to take a photo of a building square on – how many of your photos look like that? Have you even ever inspected the corners of your photos and thought them to be soft anyway? And let’s also not forget the fact that depth of field often means the foreground and/or outer edges of an image can be out of focus and therefore not worth rendering perfectly in the scan anyway…

Choosing a lens for you

Hopefully the above is food for thought when it comes to thinking about choosing a lens for the benefit of sharpness right into the corners? Actually, of course, there is much more to the process of choosing a lens than just looking at corner sharpness. In fact, Richard’s comparison touches on something he talks about in the first forum thread I linked to, as well as one of the other variables I mentioned above. That being lens compatibility and ease of use on your camera.

The Vivitar lens he used is actually an enlarger lens and therefore a lot less easy to mount onto a camera than the Sigma lens. The Sigma lens might be a lot more expensive, but it’s also a lot more convenient for cameras that it will mount directly to.

At this stage, the question of cost vs. convenience vs. quality comes into play. That is to say, in deciding what lens is right for you, you don’t just need to think about the quality of the output, you also need to consider your budget and how willing you are to faff with lens adapters, bellows, etc.

It’s also fair to say that there is no rule as to what type of lens is “best”. I.e. just as there are some expensive lenses that are easier to mount to a lot of different cameras, some of the best lenses are very hard to mount. It’s also true that whilst most of the best lenses are expensive, some expensive lenses aren’t the best for this purpose. As Richard notes in the forum thread, “I’m surprised that some high-pedigree lenses (Canon 100 L Macro, Zeiss 60mm f/2.8 Macro C/Y, 100 Bellows Elmar, etc.) show poorly in these demanding tests at 1x. I’ll assume that these lenses perform well at other magnifications. Performance at 1x is a tough standard for lenses.”

…And all this is even before we get to the availability of some lenses on the new and second-hand marketplace.

You can perhaps now see why I have previously found this topic a difficult one to tackle…? Coming up with a definitive answer as to which lens is best is very difficult, and indeed why ultimately that question comes down to the needs and selection criteria of the individual photographer.

Richard’s lens lists

It’s also why Richard’s lens lists are really quite useful. With everything I have so far discussed, hopefully, it will now be a little easier to establish just how good a lens you need for your digitisation use case. And with that knowledge, combined with Richard’s lens lists you should be able to more easily find a lens that meets your need. Of course, I it should be acknowledged that Richard hasn’t tested every lens out there. It should also be accepted that copy variance exists in lenses. That is to say, there might be “better” lenses out there, and indeed even if you do buy one of these lenses you might not get as good a copy as one that Richard tested. That said these lists represent as a good place to start as any I know of when looking for a lens for this purpose… And if you are really concerned, once you have bought the lens, you can always get yourself a Vlad’s test target and test it for yourself.

On to the lists:

Tier 1 lenses

In Richard’s words, “Testing done at 1x magnification to 50MP Sensor, mid-corner, from the Edmund Scientific glass USAF Test target (which has perfect lines to 224 lp/mm)”. Testing this way, these are the lenses that performed best:
50 f/2.8 Sigma EX DG macro
50 f/4.5 Tominon
55 f/2.8 Micro-Nikkor AI or AIS
70 f/2.8 Sigma Macro ART
75 f/4 APO Rodagon-D 1x Copy Lens
75 f/4.5 APO Rodagon-D 2x Copy Lens (normal orientation)
90 f/2.8 Sony Macro
100 f/2.x CoolScan 8000/9000 lens

Tier 2 lenses

These are lenses Richard has found to be “…nearly as good but have slightly less contrast or resolution”. They will give results that are hard to distinguish from tier 1 lenses.
35 f/4.5 Tominon
50 f/3.5 Canon FD Macro
60 f/2.8 Micro-Nikkor AF
75 f/4.5 Enlarging Ektar
80 f/4 Olympus Bellows Macro
80 f/4 APO Rodagon-N
80 f/5.6 Rodagon, fixed aperture lens
80 f/5.6 Componon
80 f/4 Componon-S
90 f/2.8 Tamron SP Macro
90 f/2.5 Vivitar Macro w/Vivitar 1x Extender
105 f/2.8 Micro-Nikkor AF-D
(I actually use the bottom one, the 105 f/2.8 Micro-Nikkor AF-D myself – I have never been anything other than very happy with the results.)

Tier 3 lenses

Richard says these lenses are “Still good”. He adds that you “will be able to see differences on real images in large prints or in pixel-peeping.” It’s also worth pointing out that when used at lower magnification (digitising larger film or on APS body), results will be closer to the lenses above.
50 f/1.8 Olympus Zuiko (on Vivitar Macro Focusing Teleconverter)
50 f/2.8 El-Nikkor (Old style, metal body, huge numerals)
50 f/2.8 Componon-S
50 f/3.5 Olympus Zuiko Macro
50 f/4 Componon
50 f/4 El-Nikkor
50 f/4.5 Enlarging Ektar
55 f/2.8 Industar
55 f/2.8 Vivitar Macro (also branded as Panagor, etc.)
55 f/2.8 Yashica ML Macro
55 f/3.5 Micro Nikkor (earliest compensating aperture type)
55 f/3.5 Micro Nikkor (later versions)
75 f/4 El-Nikkor
75 f/4.5 Enlarging Ektar
75 f/4.5 Tominon
90 f/5.6 Komura (an enlarging lens)
100 f/4 Canon FD Macro
100 f/2.8 Canon EF L Macro on 5DSR or on Sony
100 f/4 Pentax Bellows Lens
100 f/4 Leica Bellows Elmar

Tier 4 lenses

With these lenses, you, “…will see differences vs the above lenses in corners, especially on close inspection of prints. Some of these lenses are excellent for other applications, but not so much
for 1x camera-scanning.”
50 f/1.8 Nikkor AF or MF on Vivitar Macro Focusing Teleconverter
60 f/3.5 JML (M39, enlarging lens? )
75 f/4.5 Componar
60 f/4 Noflexar Bellows Macro
60 f/2.8 Zeiss Macro C/Y
75 f/3.5 Vivitar Flat Field Lens
80 f/5.6 Vivitar VHE
100 f/4 Minolta Bellows Lens
100 f/3.5 Yashica ML Macro
105 f/4 Bellows Nikkor
105 f/4.5 Tominon

Tier 5 lenses

These are the lenses which tested with image quality that Richard judged as unacceptable due to CA, low contrast, or loss of resolution in my mid-corner measurement.
50 f/1.8 Nikon E prime lens on extension tubes.
50 f/2 Nikkor AI prime lens on ext tube
50 f/2.8N El-Nikkor (newer version, plastic body, small numerals)
105 f/2.8 Micro-Nikkor AF
105 f/4 Noflexar

Interestingly, the first two on the list above aren’t actually macro lenses but are just normal prime lenses mounted to extension tubes to increase their magnification. This is commonly thought of as not the ideal setup, yet I know of a few pixl-latr users who use lenses like this and see no issue in their results due to the fact that they either shoot low-quality lenses or with pinhole cameras in the first place. It’s true that if pixel-peeping they might not be able to see the grain, and that the digitisation lens adds more smearing to the corner of their images. Try pointing this out to them though, you might find it interesting just how little they care… …

Some final thoughts

As I have said a few times throughout this article, the question of how good a macro lens do you really need for DSLR/mirrorless camera scanning is not a simple one to answer. Simplifying the question down to how sharp a lens is into the corners can give us some very helpful clues, and we are lucky to have people like Richard out there who are willing to spend their time and money making these comparisons as they can definitely help point us in the right direction.

With that said, I think it is even more important that this decision is approached with a little bit of pragmatism. Thinking not just about how sharp a digitisation lens is into the corners, but also how sharp we need a lens to be into the corners. Do you or your images and how they will be displayed actually require perfect levels of sharpness? If for whatever reason the answer to that question is yes, then hopefully Richard’s lens test lists prove useful to you? If the answer is less clear, then hopefully this article has made it easier for you to find the balance in your personal cost vs. convenience vs. quality equation!

Finally, I asked Richard for a summary recommendation, this is what he sent me:

– The easiest approach is to use a current AF macro lens in your camera mount, one that focuses to 1:1
– A less expensive is one of the legacy macro lenses mentioned above, with an adapter. Just bear in mind, some require an extension tube for 1x
– Less expensive still, but more trouble is one of the enlarging lenses, 70-80mm will be most convenient, with an adapter and extension tube

Thanks again to Richard for his part in this article, and thanks to you for reading! If you have any questions or experiences you would like to share with lenses you have tried, then please comment below!

Those links to forum threads again:
What is the best lens for 1x Camera-Scan?
How Great a Lens do we Need for Camera-Scanning?