I wonder, re: trichromatic light, if 16bit files & workflows help reduce the need for this sort of effort, despite its clear utility for digitizing color films. Us 14bit users definitely can use the clean data, if not the added work of the process as it stands now.
As @VladS has said, 16bit provides a lot more data to use to achieve the desired goals. Otherwise, why wouldn’t the bigger companies focused on cultural heritage work add trichomatic workflows to their high end 16bit systems?
This is exactly what the whiteClip and blackClip settings are doing, unless I misunderstand what you are asking for?
I did consider making this a slider, but ended up making it an input box to save vertical space and break up the long group of sliders.
I also considered having a single slider called “Dynamic Range” but found I most often only needed to adjust one end.
Open to feedback though!
@SSelvidge RE limited DOF of scanner lenses: Yes, I started with an old enlarger but it was too bulky and didn’t allow for any adjustment of the focal plane. So I constructed something horizontal that is very stiff. On this, the camera plus lens combo can move for- and backward on a linear bearing set driven by a fine thread (from an optical bench) to enable accurate focusing. The camera plus lens combo can also move up and down, back and forth, and sideways using rails to enable positioning of the sensor with respect to the negative (these are simply sliding adjustments and manually controlled). The negative holder can rotate on a panorama head to adjust horizontal alignment while I control tilt for vertical alignment using shims. I thought about these 3-point screw-adjustable tripod heads but I can get away with the current setup. You will need accurate adjustment of all degrees of freedom to align the negative perfectly to the sensor and the rig needs to be stiff enough to retain the alignment, also during focusing (which should be controlled and perfectly perpendicular to the sensor/film combo. Hence something from an optical bench). I align the film and sensor every time I re-mount the camera/lens combo (NEX7+minolta 5400 scanner lens) on the rig using Vlad’s target (which is brilliant!). I use focus peaking but this is not accurate enough, so I use the onset of focus peaking, looking at the edges when these just start to peak, and do this approaching focus peaking from both directions. When alignment of film and sensor is properly adjusted, then the onset of focus peaking happens at the same time in all 4 corners and for both focusing directions. During scanning I zoom in looking at that on a large monitor connected with HDMI and focus on the grain. I repeat/confirm that every couple of pictures. In short: You need a very stiff rig and accurate adjustment of all degrees of freedom to align the film with the sensor. This way, the limited DOF of scanner lenses is not an issue.
RE printing lenses: I didn’t look into these but there’s a good macrophotography website where lenses are compared (linked in my blog). My Minolta 5400 lens should be capable of around 5,000 dpi, so about 80 lp/mm or roughly 36 MP if I recall the ratios correctly. I achieve around 60 lp/mm with my 24MP NEX7 which is the limiting factor and this is plenty good for (too) sharp prints at A3 so I don’t see a point in pushing further (for me).
RE dynamic range. I didn’t even look how much bits my NEX7 is, but I suppose it’s 12 or so. A well exposed and developed (B&W) film (in my case mostly Kodak double x) will nearly completely fill the histogram while scanning, so will approach the dynamic range of the NEX7 but not reach it. I verify exposure for each shot to be just shy of clipping with 1/3 stop of exposure variation. On a good negative, the histogram is then just shy of clipping the highlights and has a little room (maybe 10% of the total histogram width) at the dark end. Hence, for me, even with the best dynamic range I sometimes get with well exposed and developed film and good film, the dynamic range of the NEX7 seems sufficient.
…if the scanner pipeline (sensor, A/D converter, amplifiers…) are low noise enough.
…help to see things that can otherwise be hidden to other wavelengths. Astrophotography often uses many different wavelengths for just that: Seeing things differently.
Using component images instead of composite images has started colour photography and mostly movies, e.g. Technicolor.
Sturdy is the thing that helps to scan repeatedly and in high quality. Components for such setups can come as generic step-up and -down rings, macro tubes or from companies providing such components for science (e.g. thorlabs) from optical companies (e.g Schneider-Kreuznach) or others. “All” one needs is an appropriate set of tubes, clamps, rails and possibly a focusing helicoid or two….and sum things that support setting the rig up. Don’t expect to small change for such parts, threaded breadboards (which could serve as a base) come at three digits for appropriate sizes. Adding a 3D printed filmholder in such a rig would be, hmm, no good idea.
Thanks for the clarification and you understood me perfectly.
I do use them like a dynamic range slider on either end of the histogram. A lot of people like the look of old prints, older films, scans, etc, and sometimes I think what they like is what IS NOT there as much as what is. And it is true that, typically, its mostly black point that benefits most from clipping.
In my head, if there were a slider it would only go in the negative direction, starting from 0, and would be paired near brightness to fine tune the desired location of the distribution on the histogram. That would be more accessible to some users visually, perhaps, and makes sense to me. You indeed already have this ability though, just split into more pieces. I am totally happy with it, but I would say I am a more advanced user. Perhaps the clipping points can be moved up nearer the brightness slider? That all said, you have bigger projects you are working on I’m sure!
Thanks!
>>> RE: “When done right, grain, highlights, and shadow loss all look natural, like a properly exposed and printed photograph from the old days. Most home scans, however, clearly look “scanned,” which largely undermines the effort of shooting film at all.”
Great point! I’ve noticed how much this directly relates to this modern trend of getting the “Film Look”. Most people mistakenly think that the look of film is “lo-fi” and can’t come close to the quality of digital. But what is really happening is that this “Film-Look” is actually more acurately the “Film-Scan-Look”. Or, even more accurately, the “Film-Lab-Scan-On-Auto-Settings-Look”. If you think about it, how many people today have actually seen a picture made with film on something other than a digital screen? Converting an analog medium to a digital one is the biggest challenge we have with modern film photography, in my opinion. For my personally, trying my best to replicate/showcase what film is really capable of as best as I can in a digital space, is my goal with scanning/editing my film photography. Seeing film in a digital space is how the vast majority of people are experiencing/sharing film photography nowadays (of course printing is still availalbe/possible, as are gallaries, etc). It is a shame because most people do not realize that the quality of film is actually INCREDIBLE and rivals (actually surpasses in some respects) digital photography. I actually shoot film not because I like the lo-fi “Film Look”, but for the very opposite reason of preferring a “hi-fi” rendering, and film helps me acheive this much more naturally/realistically than digital can 
I totally agree. Coincidently, I published another article today at film4ever.info - Color Balance in Film Is Not a Creative Choice about how people terribly mishandle Kodak VISIOIN3 500T stock. I extracted the five-color palette from Lomography public galleries filtered by film stock. Here is the result. Wonderful 500T all goes into green and blue because people believe they can “fix in post” 
