Voice speed changing method

Abstract

The invention discloses a voice speed changing method which comprises the following steps of: 1. acquiring an original voice signal and taking a frame as a processing unit of the voice signal; 2. inserting or deleting a section of voice signal at an original voice signal frame to be processed so as to realize slow-playing or fast-playing effect; 3. acquiring a largest self-correlation frame of the original voice signal frame to be processed by adopting a self-correlation method, wherein the voice signal content from the largest self-correlation frame to the original voice signal frame to be processed just is a section of voice signal inserted in a slow-playing way; and 4. acquiring a largest self-correlation frame of the original voice signal frame to be processed by adopting a self-correlation method, wherein the voice signal content from the original voice signal frame to be processed to the largest self-correlation frame is a section of voice signal deleted in a fast-playing way. The invention can realize good tone quality and speed changing effect on the basis of low computation workload, low power consumption, simple algorithm and concise soft implementation codes.

Description

technical background [0001] The invention relates to a method for changing the speed of speech, and particularly designs a method for changing the speed of speech with frame as the processing unit. Background technique [0002] Speech variable speed, or voice companding, refers to the change of speech speed during voice playback, including fast forward and slow forward; just like we speak, we can speak faster or slower; if we play fast, we can speak faster Listen to everything in less time; slow down and hear more clearly, word for word. In fact, the simple speed change of voice can be realized by changing the playback sampling rate; but the speed change by this method will seriously change the tone, and the playback effect is completely different from the original voice, such as male voice becomes like female voice, or female voice sounds like male voice. The speech speed change described in this article refers to changing the speed without changing the pitch. Voice vari...

AI Technical Summary

Problems solved by technology

The advantage of the above method is that the voice signal is processed in the smallest basic period unit for subsequent variable speed processing. The disadvantage is that the algorithm is more complicated and the code is more cumbersome when the software is implemented.

But there are some common defects at present, that is, the sound quality after processing is obviously much worse than the original sound source, there are vibrato, and even slight unknown noise is introduced, especially the slow playback effect is difficult to meet the needs of users

Another problem is that the number of stages or multiples of variable speed is limited. In some products that are usually seen, the slowest slow playback is 0.65 times the original speed, and the fastest fast playback is 1.5 times the original speed, or even if it can reach Slow playback by 0.5, fast playback by 2 times, but the effect at this time is already unbearable

Images