Usually in a radar signal-processing scheme, the primary consideration is the geometrical resolution of the system, or, if a particular measurement is desired, the accuracy of the system. Thus a given signal-to-noise ratio may be sought in order to meet a simple detection requirement or a desired accuracy. Usually, beyond these simple criteria, the matter of signal amplitude fidelity is not particularly stressed. In many processes, however, signal fidelity can be an important consideration in evaluating the over-all effect on the observer. In this paper the loss of information caused by signal interference in correlation or matched filter techniques is evaluated in an approximate manner. This is related to the invariance of the total integral of the Woodward ambiguity function. Degradation of the desired signal due to the statistical fluctuation of the signal itself is also considered. The distinction between pre-detection integration, which is involved in the correlation process, and post-detection integration is made. How these fluctuations can be reduced by post-detection integration is discussed and the resulting probability distributions of the signal are evaluated. Finally some brief comments are made on the philosophy of providing adequate gray-level rendition in radar displays. Attention is given to reconciling the dynamic ranges of the display and the signal, and to the number of resolvable gray levels available in the signal. A good example of these principles is the problem of recognition of objects in maps obtained from high-resolution radars.