On Film Scanning and Editing Ethics

1. Getting started

With the rising popularity of film photography in a mainly digital world, most people, including myself, rely on a hybrid workflow to process and show their work. Printing in the darkroom used to be the norm, and while it is still very much alive today — and worth the time and effort — it remains a costly and time-consuming process which is not practical for most photographers. 

Film scanning has become the new norm for film photographers in the 21st century. When I was just starting to get interested in film photography, the problem of how to go from a piece of film to a final image arose rapidly, and it became clear that I couldn’t afford to keep printing on silver gelatin paper with my old Durst enlarger forever. Scanning, on the other hand, allowed me to easily upload my photographs onto social media and internet galleries. 

But where to start? 

Information regarding film scanning has gone from scarce to overwhelming over the past few years, and in this plethora of information pointless controversy, misconceptions and pecuniary interests have flourished; from overly technical debates on scanning systems to the promotion of professional scanning solutions none of us, regular folks, will ever get our hands on; from dubious YouTube reviews to brands advertising 600€ 3D-printed film holders; from plug-in makers and film presets it has become easy to get the wrong idea about film scanning — namely, that it can either be done on 30€ scanners bought from Amazon, or by investing thousands in exclusive devices and software. 

The aim of this article is simply to explore a third and, I hope, more reasonable path: I think that it is entirely possible to get remarkably good results with a moderate investment, some patience and a little bit of DIY.

2. A Reasonable Set-Up

Some time after I started to experiment with film, I inherited an old and somewhat foggy Epson V700. For those who don’t know, it’s a flatbed scanner with a dedicated film-scanning set-up, and while it is a very capable machine with the right piece of software (VueScan being, in my opinion, the best option) the foggy glass induced such a loss in sharpness and contrast that I grew frustrated with the results.

When I was able to replace this scanner with something else, I decided to go for a digital camera and a light table instead. The reason for this choice was the flexibility it offered: I didn’t want to only scan film, I also wanted to be able to carry a digital camera with me and do some video work; flatbed scanners are a good option, but they won’t allow you to capture images of your friend’s birthday when you’re not scanning film with them.

As I mentioned earlier, there is a lot of debates going on about which camera system is best-suited for film scanning with a lot of people deeming anything below Fujifilm’s GFX line is not even worth mentioning. I want to suggest here that pretty much any decent camera body will do perfectly fine so long as you can pair it with the right lens, and by that I mean that you can definitely scan and get beautiful results with a 2012 EOS 6D, which you can get for practically nothing. The smaller sensor and lower resolution might seem like a deal breaker, but they really are not as I’ll try to demonstrate. 

In my case, I settled for something a little more modern: a Fujifilm X-T3, which I bought a short time before Fujifilm released the X-T5. What I really needed was a lens. From the information I was able to gather at the time, most people recommended macro lenses. As I couldn’t afford most of them and was still unsure about it, I needed an alternative solution. 

A good scanning lens must have a few specific properties. Firstly, it must have a flat field of view — or as flat as it can get. Secondly, the brightness and sharpness must remain constant from edge to edge. This is the exact description of an enlarger lens, and for about 50€ I was able to get my hands on an EL-Nikkor 50/2.8.

The rest of the set is barely worth mentioning: a light table, a few extension tubes,  a Fuji X to M39 adapter, and a tripod. Add in a few home-made cardboard masks for the DIY part, and you’re good to go.

Now, my actual scanning workflow is very simple: I sandwich my strip of film between the two parts of my cardboard film holder, adjust the focus by moving my camera up and down the tripod until the grain is absolutely sharp, capture the images, and upload them to my computer. 

The sandwiching part is important because it allows for the holder to lift the film above the light table surface, saving you from ring-shaped aberrations also known as Newton’s rings.

3. Workflow

Disclaimer: this is not (really) a tutorial.

I typically like to work with high-resolution images, and while I think it is pretty useless to try and get 100MP out of a piece of 35mm film due to the film’s inherent resolution, a scan in the 30-50MP range suits my needs and tastes. To reach 50MP with a 26MP APS-C sensor, I take a series of three images which I then stitch together in Affinity Photo. The algorithm is generally good enough, but you’ll want to make sure it did not create halos on the stitching line — which usually happens around the edges of the negative, or with thin, elongated shapes such as wires.

The stitching process can be done with a variety of tools, I use Affinity Photo but the same can be done with Photoshop or a free panorama tool like Hugin.

Now comes the important part: inverting and editing the scan to produce a final image. As I said before, there are a number of tools being advertised to do this. Some like Negative Lab Pro or the inversion algorithms of flatbed scanners’ software pretty much automate the whole process but don’t let this fool you: it is absolutely possible to do this all by hand. My tool of choice is Darktable. It is a free, open-source, multi-platform software that will let you organise your photo collection, as well as edit and export your images with an impressive collection of tools called Modules

Darktable’s modules are on the right-hand side. The most important one for us is called Negadoctor. It is a dedicated film inversion tool which lets you sample the negative base and perform colour and contrast corrections. For a long time, I only performed basic inversion with Negadoctor, leaving fine-tuning, white balance, and contrast control to other modules. I would also used masking, for burning and dodging purposes, extensively. I recently found I could perform most of these tasks within Negadoctor, and save curves and specialised modules for particular cases. The result in my workflow was a sharp decrease in my module stack’s size, and faster edits. As demonstrated in the images below, I now can get the job done with a simple sequence of:

  • Inversion
  • Colour correction
  • Exposure calibration
  • Colour calibration
  • Contrast and chroma tweaks

This is, of course, just a quick example, and if you compare this image sequence to the final image, I did a little more than that (but not much, really): some fine-tuning, an RGB Curve module and that’s it.

4. On film editing

It is an understatement to say that scanning and editing film, especially colour negatives, can be a controversial topic — I worked in a lab, trust me on this. While film stocks undeniably have their own set of characteristics, from grain structure to spectral response, your negative image is first and foremost a starting point or, to quote Ansel Adams from memory, a score waiting to be played and interpreted. 

Back in the days of darkroom printing, everything from your lens to your enlarger’s filter settings and paper of choice would influence the look of your image, and I’m not even mentioning darkroom tricks. What we are doing on computers nowadays is not so different from what used to be done by hand; the only difference is the degree of precision we can achieve, and the ease with which we can alter the final look of a photograph. In the age of image manipulation, forgery and AI-generated slop, altering your scan can feel like cheating, or robbing the film from its characteristics, but it really is not, and I am certain any reasonable person knows where to draw the line between editing and fabricating an image.

When I am increasing the contrast of a photograph, or choosing to push the colour balance beyond what my eye saw to emphasise what I felt at the moment, I am editing a photograph; if I am removing or adding elements or persons, replacing the sky, liquifying a silhouette and whatnot, I am altering the image, fabricating a reality that never existed and thus misleading the viewer. 

I think it is important to adhere to a certain ethic when it comes to photography in the age of disinformation and AI-generated content. I set my limit in editing to what can be done in the darkroom: anything, regardless of its technical complexity, which can be done in the darkroom is permitted on the computer.

5. Editing ethics

But mentioning technique only is not enough. I love photographing people, and while I will always refuse to alter someone’s features to transform them into someone they are not — something that I make very clear every time I photograph someone — the question of how far to extend this principle is a central one. Photography and reality have a complex and troubled relation to one another: a photograph captures reality as much as it twists it; a photograph is never the truth, but it is not necessarily a complete lie either. 

Every time I’m sitting at my desk with someone else’s image on the screen, I have to solve a dilemma: how far is too far? As I mentioned, there are things I refuse to do: I won’t transform anyone into someone else through digital manipulation, but what about blemishes, or pimples? That vein that pops a little too blue under the skin, this lone hair the flash’s shadow made too obvious, or those wrinkles you only notice on a still image?

I would love to be able to say I have never removed any of these, but it wouldn’t be the truth. The rule I work with is this: never add anything that wasn’t there in the first place, never remove something that is essential to your subject’s identity. Or to put it a little more lightly: pimples are temporary, your freckles are forever.

ML

Using Format