Subjective Evaluation of Transmission Delay in Telephone Conversations

An earlier experiment by Riesz and Klemmer on the effect of puretransmission delay upon natural telephone conversations was extended in a test with more than double the time period and number of calls. The previous finding of little or no adverse reaction to round-trip pure delays of 600 and 1200 msec alone was confirmed. The previous finding of a large increase in dissatisfaction with both of these delays following exposure to 2J+00 msec was not obtained. Exposure to delays of 2J+00 msec led to no dissatisfaction with later calls at 600 msec, but some rejections at 1200 msec did occur. There is no contradiction of other results on normal telephone circuits with 2-wire terminations (and related echo sources, paths, and suppressors) wherein customer dissatisfaction is greater with 600 msec delay than with the much shorter delay of a normal long-distance circuit. A previous paper by Riesz and Klemmer 1 in this journal described laboratory experiments on the effect of transmission delay upon the quality of telephone circuits for normal conversation. These experiments were of two types: (i) "pure delay" tests in which long transmission times were employed, but the side effects of echo and echo suppressors were avoided by using special 4-wire telephone circuits and (u) "2-wire" tests which used long transmission times in normal 2-wire circuits (or circuits with 2-wire terminations) with echo sources and echo suppressors. Since the publication of these experiments, several evaluations outside the laboratory have been done on circuits with long transmission times and naturally-occurring telephone calls (e.g., Helder,2 Klemmer 3 ).