Scanning Film with a Digital Camera
At one time the measure of a good negative was the ease or difficulty it provided in making a paper print with an enlarger. Sadly the technology for creating analog prints from color reversal appear to have been lost forever. I still value a B&W negative that prints easily, but in most cases ease of scanning is even more important because the digital representation is the first and maybe only way that other people will see the photograph.
Part of the intrigue of film is that any given piece of film contains a landscape of subtle detail. Part of the reason shooting film is such an odd choice is that for the most part, this detail is inaccessible.
The limitations of consumer grade scanning techniques results that the digital representations are necessarily compromised. When you browse the web page of any good film photographer you will notice the deficiencies in the scans—if you're looking for them. Does this matter? I doubt someone would like their own portrait more if only it were sharper!
For 35mm and medium format I prefer a digital camera over a consumer-grade scanner for several reasons: speed and control over exposure.
Always use the self-timer or remote to get to get a sharp image. Manual focus works fine, but on mirrorless contrast-based autofocus has served me well. From experimentation it seems that the best depth of field and sharpest on-center for APS-C sensors is around f/8. Keep in mind that at close range (a magnification factor of 1:1.5) the effective aperture is approximately f/11.
Digitizing film is a very demanding task for a lens. For my Fuji, I can confirm that Ken Rockwell was correct when he identified the Fuji XF 60mm f/2.4 ASPH as "optically just about perfect".
In my experience, the first edit is never your final adjustment. Walk away from the computer and come back an hour or a day later. In a way, you are partially blind while making an initial composite. With the passing of time you will almost certainly perceive each frame (and the collection as a whole) differently.
I generally keep Gimp files for two weeks, if I'm still happy with them I
export them as a
and delete the intermediate files.
Black and white negatives capture a relatively wide tonal range which does not map the linearly to the digital camera's dynamic range. This means that you will almost want to adjust the shadows and highlights using an s-curve.
Another technique, and one that I use quite often is to use the Grain Merge blend mode to simultaneously deepen the shadows and brighten the highlights. I then use the +1 layer to restore some information lost in the shadows.
Another technique you may want to try is applying a curve to all three layers, and then make final adjustments using opacity. Combining multiple images will not suffer from banding.
The range of tonal values in a color transparency is very wide, and has a vast landscape of shadow detail. Because of this I always bracket exposure. There are many complex ways to combine the three exposures using luminosity masks. but I will list the most strait-forward technique I know of.
The concept is to pick up some of the shadow detail by adjusting the opacity of layer "0" over layer "+1", then restore color and detail in the bright areas by merging the underexposed frame using a layer mask.
Mask from image on
Alternatively I may use this arrangement to create more contrast
|0||Normal||100%||Mask from image|
|-1||Grain Merge||20%||Mask from image|
One major advanage of color reversal is that you already have a perfect reference for scanning for adjustments. Keep the film the light table and tweak the image until it looks very close to what your eye can see.
After combining layers clip hilights using levels or curves. Pure white is an accurate representation for transparent areas on the film.
Target Print Size
For those occasions when I have the opportunity to make a larger print, I simply get an epic drum scan, but most images I take will be printed and displayed at a small size. Hence the goal of digitization is to make a good on-screen preview and a good 5x7 print.
|4000||Image from a 24MP digital camera|
|1900||Reduce by 50% to erase aliasing |
|1600||Optionally adjust framing with up to 20% crop |
- The image from a digital cameras most CCD scanners are compromised by a anti-aliasing. Scaling an image down is a primitive technique to force pixels to snap to more distinct values
- Most compositions benefit from some tweaks to the framing. This was the major motive for building image-review
At 300ppi this final vertical resolution works out nicely for a 5x7 on screen or in print.