Gross fog (B+F) is read from a "clear" area of a control strip; that is, an area that does not receive exposure. All films have a gross fog density, resulting from several factors that may include the following:

The density of the film base

Chemical fog

Age fog

The development of unexposed silver halides

*Inadequate fixation (film not cleared)

As stated earlier, the amount of information you use to monitor or control your process depends on several factors. However, when you choose to monitor more than one processing variable, you should construct the appropriate control chart or use a piece of graph paper that can be posted near the process. Figure 2-14 shows a typical family of control charts for a process. A family of control charts, such as this, will provide you with a wealth of information about the process. Also, all the information is in one place.

LIMIT LINES

The upper- and lower-limit lines on a control chart are based on the assumption that the plotted points are representative of a normal "population" or set of circumstances of the process. The limit lines, therefore, should include between them, all points representing an unchanged or normal process. Limit lines can never be placed in such a manner that all data are included between them; there will always be deviations. Samples from a black-and-white process, for example, show a gamma average of 0.70. On a subsequent test, a sensitometric strip was found to have a gamma of 0.80. Obviously this process appears to have changed or is changing. Should the process be altered? The answer must consider the factor of probability.

Two risks are involved in judging whether normal limits are exceeded. One risk occurs when a certain sampling appears outside one of the limit lines, indicating that the process is out of control, but the process is actually behaving normally and has not changed. This situation is known as the alpha risk. The reverse is also possible; it appears that the process is normal when actually it has changed or is changing. This is called a beta risk. These occurrences cannot be eliminated, but they can be reduced to the point where the probability of their happening is small. One risk is usually more costly than the other, and the limits are set accordingly. The limits are set far from the mean when the alpha risk must be avoided. They are set close to the mean when the beta risk must be avoided.

It is standard practice in black-and-white processing to place the limit lines at three times the standard deviation above and below the mean, or 3s. The alpha risk is approximately 3 in 1,000 for limits of 3s. Before proceeding, it is necessary to define the following two terms:

Population-all possible results (happenings) in a certain process

Variability-the amount of departure of measurements (parts of the population) from the mean (average)

Variability may be expressed in the following ways:

Figure 2-14.-Control chart examples.

Range-the difference between the largest and smallest numbers of the set

Deviation-the amount each element of the population is away from the mean

Standard error-the average percent of deviation from the mean