Because K7 is designed to separate the gray values between 0 and 255, you should calibrate your imaging techniques to a dMin of 255 and dMax of 0. As elementary as this sounds, many users mostly automate the use of their scanners and cameras, and their Photoshop workflows without regard to controlling the process.
Turn off scanner automation in the software that you use to run your scanner.
Some of the options in Preferences as an example, will automatically perform a type of “levels” to make the lightest pixel a white point, and the darkest pixel a black point. While this might be handy for making a snapshot of a polar bear in a snow storm, it is not in the best interests of a photographer to force any and all light situations into absolutes. The goal is to get the scanner to accurately capture the density present in the negative without modifying it.
Scan a negative as a positive.
I tend to scan my negatives as positives so that I can think of the values of the negatives without having to reverse my thinking. Also, some scanners use an algorithm to convert from negative to positive when the negative option is used. I want to avoid that. The most important part of scanning is to create a white point and a black point. That white point and black point are definitely not always the lightest and darkest portions of a negative. Instead I create a set of settings using a particular film type or perhaps a film type/development process so that each time I scan this type of film I can select the settings and then I can have an automated process.
For the white point and black point settings, I focus the scanner white point onto film base + fog for the 255 value (white point), and I focus some bit of exposed film leader as the 0 value (black point). The resultant scan need only be inverted in Photoshop to appear as a positive, and will now contain all the possible exposed tonal values in relation to this film type. A specular will reproduce as paper white. That part of the film that did not receive any light will reproduce as black. Everything in between will fall into a natural range with the following caveat.
The mid-point of a negative needs to be arrived at as a L value of 127.
If the photographer can identify that part of the negative through their exposure log; or their eye; or by taking a picture of a 50% gray card (not an 18% gray card), the gamma level during scanning can be fixed so that this portion of the film produces an L value of 127. Sometimes the mid-point tool will push the value to 127. Sometimes the user must slide a Gamma slider to adjust this value so that it scans as 127.
Save your settings
At this point the scanner is set up to scan this particular film type, and all subsequent scans of this film type will produce a normalized input scan when these settings are used.
How can one really “know” the 50% gray of a film type?
If the film is still available it is possible to expose diffused white light in successive 1/2 stops and develop the film to arrive at the densitometric measurement equal to the mid-point. That particular frame, with the film base + fog, and a fully exposed leader is all that is necessary to calibrate a scanner to properly scan a negative. If these three points can be captured to L values 127, 255 and 0 when scanned as a positive and inverted in Photoshop, all the other gray values will fall into their natural place.