Voice processing method of portable Chinese tone learning machine

Abstract

The invention discloses a voice processing method of a portable Chinese tone learning machine, wherein the learning machine not only has announcer (standard) pronunciation to be repeatedly read but also can immediately display the tone curve of the repeatedly read tone. The learning machine simultaneously displays the repeatedly read and the standard two tone curves, has the difference of characters promote quantification, and leads customers to immediately know the improved direction and effect, so as to help the customers to improve learning efficiency. The voice processing method consists of three parts: the preprocessing of a signal, comprising the steps of defining the voice characteristic of the customers and preparing before processing data. The signal is analyzed and a display result is processed with a band selection analyzing method. The voice processing method is designed for a brief voice signal. After changing a process work flow and a preset parameter, the method is applicable to adjusting the melody of a piano and judging parameters such as tempo, tone, generated tempo frequency, and the like, with extremely high precision (equal to or more than 0.1cent). Definitely,the method is also applicable to analyzing or measuring the pitch, the tempo, and the like, of the other musical instruments.

Description

technical field [0001] The invention is the core technology of the Chinese tone learning machine. A technical field dealing with speech analysis. In particular, it can be used by various groups of people to learn Chinese tones, and can accurately extract the fundamental frequency value of each pronunciation and give a tone curve. This technical feature also belongs to the technical field of phonetics. Background technique [0002] Chinese is a tonal language. This is also one of the outstanding features that distinguish it from Indo-European languages. Tone is also a big difficulty that people encounter when learning Chinese. To conquer it, the only way is to practice pronunciation repeatedly, there is no other shortcut. Anyone of any nationality learns to speak by repeatedly imitating the voice of adults countless times. No one can escape this learning law, and people spend a long time learning it. However, adults do not have the time and conditions for children to lea...

AI Technical Summary

Problems solved by technology

One of the reasons for the low accuracy of this frequency extraction technology is that it is limited by the uncertainty principle, which has been recognized by some users (see Document 2)

Images