In order to describe acousto-optical spectrum analyzers (AOSA) it is necessary to take into consideration not only spatial resolving power defining the amount of the spectrum intervals to be distinguished (which is usually considered) but also data temporal resolving power for temporally variable spectra. These parameters are intercepted, and changing of one influences both on variation of the others and on the AOSA total information efficiency (amount of useful information extracted by the device from the variable signal of certain duration). It has been shown that increasing of the signal transmission through the acousto- optic Bragg cell exercises influence on the AOSA information efficiency by the complicated way. The bigger this time, the bigger amount of the spectrum resolved intervals (which means that the information capacity becomes bigger), the less the signal useful duration, and the less the AOSA information transmission capability for the devices processing the signal with periodically variable spectrum. Along with that part of the time is spent for mixed distorted signal. In order to optimize AOSA from the point of view of useful information transmission, it has been proposed to assign different weight factors to each of kinds of information, and, correspondingly, to each of the device parameters. These weight factors define valuability of certain kind of information. Their using for the quantitative estimation of the useful information amount, allows to create the basis of AOSA information metrology.