(Go back one page to co-stained compensation samples,
or back to the introduction).
Below is a figure showing a sample of peripheral blood lymphocytes stained with PE, and collected at several different compensation values (vs. FITC). The top left panel is uncompensated; top right is under-compensated, the bottom right is over-compensated, and the bottem left is properly compensated. The bold yellow line has been added for emphasis only; no quantitative relationship is implied by the curved shape of the line! In this example, the sample has been gated for lymphocytes already, so that the autofluorescent monocytes do not interfere with the analysis.
When compensating, you should make an attempt to visualize the equivalent
of this yellow line. Mentally line up the centers of the populations. The
ideal is that the positive and negative populations line on a straight line:
that is perfect compensation. The line should be straight for the entire
range of cells: from dimmest to brightest.
Note in the slightly-undercompensated panel, you might think that the compensation is correct when you compare the main population of stained cells against the unstained--it is quite close. However, the small population of very bright cells is still clearly undercompensated. This is a classic example of why you should use the brightest stain possible for compensation (for a fuller discussion, see Requirements for Proper Compensation).
In the slightly-over compensated panel, the same effect is seen. This example is a little more insidious, because the overcompensate bright population is against the left axis, and (in a dot plot mode), might be easily missed. Again, the dimmer positive population still looks properly compensated.
Go on to an example of improper 3-color compensation.