Here is a white paper I wrote a couple of years ago. My rig has changed a bit, but the conception is the same:
As far as calibration is concerned, that all starts with choosing a good monitor MADE FOR graphic arts and photography. It continues with proper ICC color management, achieved through the use of a color calibration kit appropriate for your monitor. Once you have a properly calibrated monitor, you can make the correct settings in your software to take advantage of it (if applicable). In Lightroom Classic, that would be in the printing module if you are printing directly to an attached local printer.
As far as anything else, as an ex pro portrait lab production manager who ran color correction and implemented color management, I’ll say that working with many different film stocks is a challenge. Every brand, type, age, and specific emulsion number of film is going to have a different color balance.
Beyond that, film development varies a little at most labs (good labs stay within control limits). Home development is seldom as consistent as commercial development, unless you’re a stickler for agitation and temperature consistency, chemical freshness, etc.
The net result is that “Correct” or “perfect” color balance is elusive! What I strive for is REALISTIC color balance, or color balance that is believable for the scene or for my intended effect.
If you are working with ONE film stock all the time, you can always expose a ColorChecker chart under controlled lighting conditions (several different scenarios such as daylight, incandescent, 2700K CFL, 2700K LED, 4100K Cool White Fluorescents, your particular flash, 5000K CFL and LED…), and find settings in NLP that seem to be good starting points. You will quickly learn that only daylight and incandescent are really close matches to reality, because the spectra of flash, LED, and fluorescent sources are discontinuous. But you will also have some idea of what to try first.
There really isn’t any substitute for experience in color correction. Working from a standard scene (I like a Colorchecker and someone I know, photographed in the same frame and light) is a good way to get started. Be sure to do an exposure ring-around at –2 to +3 stops, because exposure slope changes color balance! In other words, you may need different color adjustments for the same subject at –1 and +1.5 stops, for example.
The hardest negatives I’ve worked with are those from the 1960s in my personal family collection. There’s not much left in them, because the dyes have faded unevenly, but some of them have made good digital files. Judicious masking in Photoshop has helped some of them.
One more thing: A lot of folks on this site seem to believe that pro photo lab scans are some sort of important reference. They CAN be, but if you are comparing production lab scans to your own scans, you should know that lab operators are doing basically the same interpretive things you are! They might have more experience, and a better scanner light source, but otherwise, they’re still humans turning knobs to correct what the machine is trying to do. In some cases, the process is automated, and you’re seeing the “personality” of the machine, not the true color of the original scene.
That is why I say, “Tune for individual preference.” Often, I’ve been able to capture better color and/or tonality from NLP than were present in the original lab scans or prints.