Pay-off between quantizing distortion and injected circuit noise

Comprehensive and generally-accepted information is available on the effects that circuit noise has on the opinions of telephone customers. With the introduction of digital technology, it is now necessary to have equally reliable information on the effects of quantizing distortion. This is being assembled largely by seeking information on the "pay-off" between quantizing distortion, represented as "speech-correlated" noise, and random, i.e. uncorrelated, noise. The required relationship can be sought by using various criteria for equivalence and several sets of experimental data are available each having fairly good internal consistency. Unfortunately, however, there are considerable differences between the sets which cannot be attributed only to differences in criteria. It is concluded that more tests are needed to explain the differences.