Avoid crackling at high speeds.

[m1maker]

Introduce a new algorithm to prevent audio crackling at high speech rates.

*by adding a conditional branch when the timing rate factor exceeds

**DECRACKLING_RATE_THRESHOLD.

*For mixed and transitional sounds (j < 132) at high rates, it processes

**samples differently to avoid crackling

However, speech using this algorithm only sounds good at high speeds, so the threshold is set.

[m1maker]

Sorry, I didn't test it thoroughly enough. When lowering the pitch, the pitch doesn't lower below the default pitch in the voice samples. Because of this, the voice loses intonation. I'll definitely fix this a little later. I apologize for the inattention. *Already fixed!