Voice Processing Method for Natural Pitch Transitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice processing technologies fail to generate acoustically natural voice signals when acoustic characteristics such as fundamental frequency change over time, particularly in transitions between steady periods with different pitches.
Innovation Solution
A voice processing method that compresses forward the first steady period and extends the transition period between steady periods in a voice signal, while emphasizing the variation in fundamental frequency within the transition period, using a computer system with a signal analysis unit and an adjustment processing unit to adjust the voice signal accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If harmonic components are moved in frequency domain to convert voice characteristics, then voice quality conversion is achieved, but acoustic naturalness deteriorates in sections where fundamental frequency changes with time
Solution Approach 1:
The voice signal is segmented into multiple steady periods based on acoustic characteristics. Each steady period is processed independently through frequency domain transformation, while transition periods between steady periods are preserved to maintain natural pitch transitions. This segmentation allows voice quality conversion while preserving acoustic naturalness in each segment and their transitions.
Solution Approach 2:
Different processing strategies are applied to different parts of the voice signal. Steady periods undergo harmonic component movement for voice quality conversion, while transition periods are preserved with minimal processing to maintain natural pitch transitions. This local differentiation resolves the contradiction between voice quality conversion and acoustic naturalness.
2Adaptability or versatility
If pitch changes are implemented between steady periods, then voice expression is enhanced, but acoustic naturalness deteriorates during pitch transitions
Solution Approach 1:
Pitch changes are prepared in advance within each steady period by adjusting harmonic components before the transition period begins. The transition period itself is preserved with minimal processing to allow natural pitch evolution. This preliminary preparation of pitch changes maintains both voice expression and acoustic naturalness during transitions.
Solution Approach 2:
The processing approach dynamically adapts to different signal segments. Steady periods receive processing for voice expression enhancement, while transition periods receive reduced processing to preserve natural dynamics. This dynamic processing strategy maintains acoustic naturalness during pitch transitions while still achieving voice expression goals.
Data Source
AI summary
A voice processing method realized by a computer includes compressing forward a first steady period of a plurality of steady periods in a voice signal representing voice, and extending forward a transition period between the first steady period and a second steady period of the plurality of steady periods in the voice signal. Each of the plurality of steady periods is a period in which acoustic characteristics are temporally stable. The second steady period is a period immediately after the first steady period and has a pitch that is different from a pitch of the first steady period.