Effects of Transmission Delay on Conversational Behavior on Echo-Free Telephone Circuits

l.i Purpose of Study This paper describes experiments that examined effects on the conversational behavior of subjects who talked on telephone circuits containing round-trip transmission delays of 600 ms and 1200 ms, typical of those in one- and two-hop synchronous satellites. However, satellite circuits also contain echoes and echo suppressors, which create added difficulties in conversation. The present study examines behavior on 115 116 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1971 echo-free circuits, which presumably represent the circuits that would result if echoes could be eliminated from existing circuits. In particular, answers are sought to the following questions: (i) Does the introduction of delay on a circuit produce any measurable effects on speaking behavior, even though the subjects notice nothing unusual about the circuit? (ii) Do the subjects experience difficulty in talking over delay circuits, even though they do not notice the difficulty? This paper reports quantitative measures of the effects of delay indicating that the answer to both of the above questions is "yes." 1.2 Background For several years, studies have been directed toward measuring customer acceptability of circuits containing both transmission delay and echo suppressors. For example, these circuits were used in one study, 1 in which hundreds of customer interviews were taken just after the customers placed transatlantic calls over such a circuit. A delay of 600 ms (as opposed to the cable transmission round-trip delay of 90 ms) produced a significant increase in the number of customers who reported having difficulty, and in addition produced an increase in other complaints and comments such as "cutting on and off," and "fading." However, such customer interview studies have limited application to echo-free delay circuits because (i) the echo suppressors are widely believed to contribute most of the degradation and probably overwhelm the delay effects, and (ii) as will be shown here, many of the delay effects are subtle, unnoticed by the conversants, and hence are not reported in interviews.