Scanning films with a digital camera
Initial version created November 19, 2020.
Updated 4 January 2021 at 05:16pm.
A - GENERAL
INTRODUCTIONI have developed my prints from film negatives for a long time with an enlarger in the darkroom, then the advent of digital in the 1990s allowed me to test other digitization or printing techniques.
In any case these methods are perfect, they just allow me to express myself, and achieve the vision I have of my work.
The difficulty is not to scan a film, but to find your method and to adapt your technique to your needs, means and above all your creativity.
Based on this learning, this guide will hopefully be useful to other people.
Depending on the purpose we want to achieve (print or print size, reproduction and conservation, use on the web, family memory) scanning (or digitization) will only be one step in the image production chain.
The scanning of a film is quite simple, you just have to convert a tiny area of the film into a series of pixels, and like under an enlarger, this requires a light source, a film holder, a lens and not paper but a digital sensor.
All scanners (Epson, Braun, Plustek, Imacon, Reflecta, Heidelberg, Canon, Optronics the missing Hasselblad Flextight and Nikon, etc.), whether flatbed, dedicated, drum, or other work somehow the same way with each their
particularity and technical quality.
The DSLR method used here is more like the work of reverse enlargement or even reproduction with a bench.
My goal using this technique is to be able to scan 135 (24x36, 35mm) and 120 (6x6) black and white negatives so that I can make prints up to 2 meters wide.
Depending on the type of film or development the grain of the film will be more or less present but as sharp as possible.
Achieving this goal is possible with rotary drum scanners, which are very expensive to acquire, require space, adjustments, preparation and maintenance.
The other option is to outsource the digitization to a specialized laboratory. Besides the cost, the distance, the delay, the lack of mastery, various tests were not more conclusive than doing it by myself.
You will also find dozens of comparisons on the net.
And finally, the freedom to be able to change the elements as I wish which is not possible on an "all in one" scanner.
120 negative film ready to be scanned by a DSLR.
B - MATERIAL REQUIRED
SOURCE OF LIGHT
I am using a Slimlite Plano LED light plate from Kaiser Fototechnik (2453).
5000 Kelvin daylight, CRI 95, variable intensity, works with battery or power supply which I recommend for more maximum and constant light.
Some take a tablet or a smartphone, or for vertical versions a third party light source or a flash. These are alternatives to explore that I have not tested.
This is almost the gist for me, and some care needs to be observed regarding the flatness of the film. In fact, the wider the film, the more it tends to curl, a phenomenon accentuated under the effect of humidity or heat.
Use a film holder that detaches the film from the surface of your light pad. Placing the film directly on the light table may result in the formation of Newton rings and create small variations of light.
I keep them flat with at least 1 ANR (Anti Newton Ring) glass underneath, specifically designed to avoid these famous Newton's rings.
I appreciated the quality of the Kaiser FilmCopy Vario holder, but unfortunately it is not possible to get the edges of the film with the "120 height".
So I opted for a Durst L-1200 Femoneg enlarger film holder with upper and lower ANR lenses, allowing me to scan films up to 4x5.
Durst Femoneg L-1200 enlarger film carrier closed.
Durst Femoneg L-1200 enlarger film carrier opened with the 2 ANR glasses.
I started with a Canon EOS 5D (12.8 megapixels), then a Canon EOS 50D (15.1 megapixels), then now a Canon EOS 5DS R (50.6 megapixels) dedicated to this job. The liveview mode is a plus to appreciate the sharpness of the grain
by magnifying and focusing on it via this feature. The control of the DSLR and the viewfinder via the screen of a computer is also a comfort allowing to obtain better results (visibility, handling, precision, fight against vibrations).
I also advise you to get a power supply AC adapter for your camera. Batteries will run down quite quickly if you leave it on permanently during scanning sessions.
Not having a Canon macro lens, I tested various lenses with extension rings or bellows in M42 mount (via an EOS EF adapter ring), including various 50 to 55 mm such as the Pancolar Electric 50 mm 1:8 from Carl Zeiss Jena DDR.
Currently, I am using an EL-Nikkor 50mm 1:4 enlarger lens from Nikon in M42 on extension rings.
Currently, I use a M42 enlarger lens Nikon EL-Nikkor 50mm 1:4 on extension tubes to assemble multiple views, beyond the 1:1 ratio.
For the 1:1, especially the 24x36, I sometimes opt for the excellent Nikon Scanner-Nikkor ED 100mm f/2.8 from a Nikon Super Coolscan LS-8000 ED film scanner.
If you want to assemble your negative from several views, you will need to use a very good lens or "macro" system, especially regarding sharpness and vignetting in the angles.
Common lenses are optimized for their own magnification limit, so pushing this factor away will give poor results, especially with blurred areas.
A reproduction bench column would probably be simpler, but I was using a 3021BPRO tripod and a 3D 029 head from Manfrotto that I already had, with the central column reversed.
I finally switched to a more accurate head to align the sensor and film with the Manfrotto MHXPRO-3WG, XPRO Geared 3 Way Pan/Tilt, allowing frame and shoot precisely by adjusting with the "micrometric" knobs.
Any head and tripod will do if they hold the unit in the starting position without flinching.
C - METHODOLOGY
SIZE OF SCANNED IMAGES
The things that affect the size of your scan are the resolution of your camera, and the ability of your optical system (lens) to magnify the film.
If you scan 24x36 format (135 film), using a 24x36 sensor and a macro lens with 1:1 magnification, your digital reproduction will have approximately the same megapixel resolution as your digital camera.
If you scan a medium format movie on a 24x36 DSLR, there will be cropping due to the different aspect ratio, and your files will be smaller in megapixels than your camera, unless you take multiple different views of the cliché that you will assemble later. It pushes the boundaries!
If your sensor has a high resolution it will be an advantage if you plan to scan for exposure prints, but for web printing and general printing any modern DSLR will do wonders.
Without going into too complex calculations, for a 6x6 image in 6 shots, with the Canon 5DS R and the appropriate lens, the result is around 16000x16000 pixels, or a file of at least 250 million pixels. If necessary,
it is possible to shoot as many takes as desired, but this often no longer makes sense, the grain becoming larger than the pixel of the sensor.
Go in the dark in order to avoid any interference of light, the complexity of the chromatic entanglements cannot indeed tolerate the interferences emanating from various light sources.
I advise you to handle your films with cotton gloves dedicated to this use.
Beforehand, clean the space where you are going to operate, as well as the lens, the light panel, the film holder and any ANR glasses.
To pass the material with an antistatic cloth, and to equip yourself with a system allowing you to regularly blow air on the area holding the film.
On this scale, you will see for yourself that the slightest vibration completely distorts the result.
Some recommend using mirrorless cameras to avoid any vibration effect when triggering, which is a good idea.
On Canon it is possible to request the permanent raising of the mirror with the Liveview, which amounts to the same thing.
If you can't shoot remotely with a computer, get a remote trigger, or use the self-timer with 2 or 3 seconds of waiting.
In any case, pay close attention to vibrations, it would be a shame to ruin everything at this stage.
The plane of the sensor and that of the film must be perfectly parallel.
The spirit level is not used much at this level, except to position parallel or perpendicular to gravity, or you need a spirit level for the lighpad and for the film plane, but this is not very accurate.
The best way is still to position a mirror on the film plane, or the light shelf and perfectly align the return of the lens image with the latter in the viewfinder.
If you want to go further, various tools like Zig-align promise to perfectly align the whole, but not having tested them I do not comment.
You can also use a negative test pattern, which will allow you to line up and possibly test different lenses for sharpness, especially at corners.
Vlad S., who is based in New York, produced a quality 1951 USAF 135 or 120 negative using the high-resolution, high contrast Adox CMS II film (see uncut film image below).
This film model is for the sole purpose of testing and fine-tuning cameras and lenses used to scan, film scanners, and to test the setup, focus and sharpness of darkroom enlargers.
You can also use the "live view" with a mirror placed on the light source and based on the lines of the screen grid by magnifying more or less. The lens image must be perfectly centered vertically and horizontally.
Find or add a very small element (dust, powder, etc ...) and place it perfectly centred on the live view grid.
Change the focus very slightly and you should get a second halo of reflection around the element.
As long as everything is not perfectly aligned in the centre, as well as the perspective projection, this is not good.
The alignment is not good, the reflection halo is not centered with the element in the live view return.
The alignment is good, the reflection halo is centered with the element in the live view return.
All of these methods will allow you to check that everything is properly aligned to obtain perfectly homogeneous scans.
The film must be positioned emulsion towards the lens. The emulsion is located on the matt side of the film, or the hollow part when it bends.
I use ANR glass to ensure perfect film flatness.
Scan done with a medium format scanner dedicated to 120 film. The negative was slightly curved, the edges are sharp, not the center.
Scan of the same curled 120 film with a digital camera between 2 ANR glasses. Everything is perfectly clean.
Without ANR glass the film is curved.
A Emulsion, B Shiny film side, C Film holder, D Light source.
With an ANR glass on top the film is flat.
A ANR Glass, B Emulsion, C Shiny film side, D Film holder, E Light source.
By using the Liveview it is possible to increase the magnification time to perfect the focus on the grain of the film.
Some people advocate autofocus with autofocus lenses, I have not tested it but I suggest doing it manually for all to ensure the final consistency.
Different Canon Liveview magnification factors up to the film grain.
Different Canon Liveview magnification factors up to the film grain.
The shots are all taken in RAW mode, the JPEG format being destructive (grain) and making stitching more difficult.
If your light panel delivers a daylight temperature, also set your DSLR's white balance to daylight.
The lens is open somewhere between f/5.6 and f/8 to have a little depth of field on the grain, and especially a maximum of sharpness and a minimum of distortion.
Lowest sensitivity possible to avoid digital noise.
From there you have to find the shutter speed that will best capture your view of the film.
Once all these parameters are set, write them down, especially if you are taking several views of the same negative, otherwise you will not get a homogeneous result.
Personally, I switch everything to manual mode to avoid any unpleasant surprises.
I am using Canon's EOS Utility directly connected to the PC as well. With the Liveview mode and the remote control of the box this reduces the complexity of handling on the small screen of the device, as well as the vibrations.
In 24x36 I take a single shot. With 6x6 film, I usually do 6 takes which are assembled later.
I tried different software such as Photoshop with Photomerge, Microsoft ICE, Hugin stitching software but without satisfactory results, until I found PTGui.
I finally bought the latter which is incredibly efficient, the negative always comes out well assembled the first time, very quickly, really impressive.
For black and white, a simple inversion, desaturation, and some adjustment in brightness and contrast will give you an immediate result.
Positive films obviously don't need to be inverted, so try the automatic adjustment as a starting point and adapt your method.
On the other hand, reversing color negative films is a complex subject that I do not practice, but you will find many articles and tools relating to this subject on the web.
D - EXAMPLE OF REALIZATION
A sample of scanning to obtain 19296x19132 pixels image (369M pixels) from a 120 negative (Kodak TRI-X 400 black and white film shots with an Holga 120 GCFN camera).
You'll probably notice on the image the reference spots added manually with a permanent marker pen on the ANR glass (negative side) to first, quickly align each shot to the next one, and second, to help automatic stitching later.
Here it's not useful because the negative subject already have a lot of control points, but when I do long exposure with large blank areas, I'm really happy to have them for both cases, believe me!
I'll later add a more specific sample with more details from scanning to printing.
ANR GLASS - HELP REQUESTED!
I'm searching for someone who can provide me specific ANR Glass cut with a grid of small numbers (around 0.5mm) printed or engraved on it.
If you have any idea or solution, please contact me
The image stitched from 24 views (4x6 files).
Final image with reference spots removed, edited for my taste.
E - REFERENCES
support and special bench for scanning with a digital camera
F - ADD A COMMENT