An increasing number of practical systems for speech communications have been proposed in the past few years. Such systems often must operate over both wideband channels and standard telephone connections. Thus, it is useful to be able to simulate the telephone channel as well as the other speech processing parts of the given system. This paper describes a digital network which provides a simple, controlled simulation of the properties of both the standard carbon microphone and the telephone transmission system. The simulation consists of a combination of nonlinear distortion, noise addition, and bandpass filtering. Both wideband and telephone signals were recorded simultaneously using a 2-channel A/D converter. The wideband signal was processed by the simulation system, whereas the telephone signal provided a reference signal for purposes of comparison. The parameters of the simulation were manually adjusted to provide optimum matches to several telephone links. The telephone simulation was then subjectively evaluated in two listening experiments. In both experiments, utterances from the simulations were paired with the corresponding telephone recording. In the first experiment, a group of listeners was asked to select the actual telephone recording from each pair of utterances. In the second experiment, a new group of listeners was asked to rank the similarity of the two utterances on a 1-to −10 scale. Results of the evaluations indicated that, for some sets of simulation parameters, the network provided a fairly good psychophysical simulation of a variety of telephone channels.
[1]
A. E. Rosenberg,et al.
Evaluation of an automatic speaker-verification system over telephone lines
,
1976,
The Bell System Technical Journal.
[2]
L. Rabiner,et al.
A digital signal processing approach to interpolation
,
1973
.
[3]
L. R. Rabiner,et al.
Evaluation of a statistical approach to voiced-unvoiced-silence analysis for telephone-quality speech
,
1977,
The Bell System Technical Journal.
[4]
Stephanie Seneff.
A Real-Time Digital Telephone Simulation on the Lincoln Digital Voice Terminal
,
1975
.
[5]
Lawrence R. Rabiner,et al.
On creating reference templates for speaker independent recognition of isolated words
,
1978
.
[6]
A.E. Rosenberg,et al.
Automatic speaker verification: A review
,
1976,
Proceedings of the IEEE.
[7]
Nuggehally Sampath Jayant,et al.
LPC analysis/Synthesis from speech inputs containing quantizing noise or additive white noise
,
1976
.
[8]
J.L. Flanagan,et al.
Computers that talk and listen: Man-machine communication by voice
,
1976,
Proceedings of the IEEE.
[9]
F. Itakura,et al.
Minimum prediction residual principle applied to speech recognition
,
1975
.
[10]
A. Oppenheim,et al.
Nonlinear filtering of multiplied and convolved signals
,
1968
.