TH29809A - Methods and devices, speech encoders and methods, and pitch detectors. - Google Patents

Abstract

DC60 (27/10/40) method and pitch detectors Even speech signals which are within half pitch or double pitch will give you a higher degree of correlation. More stable than a pitch To be detected The imported speech signal will be judged as a reverb characteristic. Or the nature of the sound that is not reverberated and the part is the sound The imported speech signals are encoded by the sinusoidal function cipher 114 and by the cipher 120, respectively, to produce the corresponding coded output. The sine wave function 114 is performed. Search a pitch on an encrypted output to find information about With the pitch from imported audio signals and a highly accurate pitch news set. Depending on the news Detectable pitch substance, results of detection Pitch captures will be decided. Using information about A sure pitch Sleep high that set like that And the results of Decision on the reverb / no-echo characteristics of a frame other than the current frame, the speech encoding method, and the speech encoding mechanism used. The pitch detection is still being revealed as well. Methods and pitch detectors capable of performing high-precision pitch detection. Even a speech signal, internally with a half pitch or even pitch, provides a more stable correlation rate than a pitch. That will be examined The imported voice signal will be judged. That is a characteristic sound Or how it sounds No echo and echo And audio parts No echo of the imported voice signal is encrypted. By the Sine Function 114 Sine Wave Coding Unit and by the Sine Function 120 Coding Unit, respectively, to produce the corresponding encoded output, the Sine Function 114 coding unit was examined. Find the pitch on the encrypted output. To find news about With the pitch from imported voice signals and a highly accurate pitch news set. Depending on the news Detectable pitch substance The results of the examination Pitch captures will be decided. Using information about A highly accurate pitch that is set like that. And the results of Decisions about echo / no sound characteristics The echo of the frame, in addition to the current framework, is a method of encoding speech. And a voice encoding mechanism that uses the The pitch detection is still being revealed as well.

Claims (1)

1. How to combine a pitch capture in a coding method. In which there is an input audio signal, separated by time axis in the form Of a pre-set encoding unit And which signal Speech that depends on the encoding unit is judged in terms of Reverb / no-echo characteristics Which contains the following components: Which is the detection of news About the pitch under the condition of detecting the correct pitch Predefined pitch and judgment procedures. Which is the determination The pitch of the current encoding unit of the speech signal. By using the effect at the top: