Electronic formant speech synthesizer

Claim 1

A SPEECH SYNTHESIZER COMPRISING A FIRST SOURCE OF VOICED SOUND ENERGY, PRODUCING VARIOUS HARMONICS AND CONTROLLABLE BOTH IN AMPLITUDE AND FUNDAMENTAL FREQUENCY, AND A SECOND SOURCE OF UNVOICED SOUND ENERGY PRODUCING A WIDE RANGE OF SUBSTANTIALLY RANDOM SOUND FREQUENCY COMPONENTS, APPLIED TO THE BASE OF A FIRST EMITTER FOLLOWER CONNECTED TRANSISTOR, A SECOND TRANSISTOR HAVING AN EMITTER LOAD AND AN AMPLITUDE CONTROL SIGNAL APPLIED TO ITS BASE, AND A THIRD COMMON BASE CONNECTED TRANSISTOR PROVIDING A COLLECTOR OUTPUT CIRCUIT, THE COLLECTOR OF SAID SECOND TRANSISTOR BEING CONNECTED ONLY TO THE EMITTERS OF BOTH SAID OTHER TRANSISTORS AS A SUBSTANTIALLY CONSTANT CURRENT EMITTER LOAD FOR SAID FIRST TRANSISTOR, CONTROLLABLE TO VARY THE EMITTER CURRENT OF SAID THIRD TRANSISTOR AND ITS GAIN, AND A FILTER SYSTEM COMPRISING A SMALL NUMBER OF CHANNELS, SOME SUPPLIED FROM ONE AND SOME FROM BOTH SAID SOURCES, EACH CHANNEL HAVING A VERY LIMITED AUDIO FREQUENCY BANDWIDTH, CONTROLLABLE BOTH IN AMPLITUDE AND IN THE PARTICULAR AUDIO FREQUENCY RANGE, HAVING A FIXED BAND FILTER IN ANOTHER FREQUENCY RANGE, A MODULATOR INPUT TO AND DEMODULATOR OUTPUT FROM SAID FILTER, AND A COMMON CARRIER OSCILLATOR FOR BOTH SAID MODULATOR AND DEMODULATOR, CONTROLLABLE THRU A RANGE TO CONVERT THE DESIRED AUDIO-PASS-BAND TO AND FROM THE FIXED BAND OF SAID FILTER, A MULTIPLE TIME DOMAIN FUNCTION GENERATOR SYSTEM HAVING MEANS TO PRODUCE A SERIES OF VARIABLE LENGTH INTERVALS, MEANS TO PRODUCE IN EACH SAID INTERVAL VARIABLE PARAMETERS CORRESPONDING TO THE SLOPES OF THE DESIRED MULTIPLE FUNCTIONS DURING SAID INTERVALS, MEANS TO COMBINE SAID PARAMETERS FOR CORRESPONDING FUNCTIONS IN SEQUENCE TO PRODUCE CONTINUOUS FUNCTIONS OF SUBSTANTIALLY RECTANGULAR STEPS, AND INTEGRATING MEANS TO PRODUCE FROM EACH SAID RECTANGULAR STEPPED FUNCTION A SUBSTANTIAL APPROXIMATION TO SAID DESIRED FUNCTIONS, SAID DESIRED FUNCTIONS CONTROLLING SAID SOURCE AND CHANNEL AMPLITUDES AND FREQUENCIES IN A MANNER CORRESPONDING TO THE PRINCIPAL COMPONENTS OF NATURAL SPEECH, AND MEANS TO PROVIDE A COMMON COMBINED OUTPUT OF THE SEVERAL CHANNELS AS SYNTHESIZED SPEECH CORRESPONDING TO SAID VARIABLE LENGTH AND OTHER PARAMETERS.