Models for the Subjective Effects of Loss, Noise, and Talker Echo on Telephone Connections

In the last 10 years, several tests have been conducted at Bell Laboratories to determine the subjective evaluation of loss, noise, and talker echo in telephone connections. The purpose of these tests was to obtain information for use in network planning studies. Several hundred volunteers served as subjects and participated in several thousand test calls with various amounts of loss, noise, and talker echo. Some of these tests were conducted on normal business calls made by Bell Laboratories employees; others were conducted in a laboratory environment. At the end of each call, the subject was asked to indicate his or her opinion of the transmission quality on a fivecategory rating scale: excellent, good, fair, poor, and unsatisfactory. The results from these tests were used to formulate the graphical and analytical models of opinion which are presented in this paper. These models have been used extensively in network planning studies 1319 to evaluate present network performance and to study the effect of possible changes in the network. Although many of these studies are not yet published, a recent paper by Spang 1 provides an excellent example of the application of these models in toll transmission planning. For many years the Bell System has used subjective tests to obtain information on the effects of transmission quality which could be used in network planning studies. For example, the present Via Net Loss design plan, which introduces loss to control talker echo in the Direct Distance Dialing (DDD) network, is based on the results of subjective tests for talker echo that are included in a 1953 paper by Huntley.