Pitch Delay Contour Adjustment for Speech Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital speech compression systems face limitations in achieving high long-term prediction gain at low bit rates due to constrained resolution and dynamic range of pitch delay adjustment parameters.
Innovation Solution
An adaptive step-size method is employed to adjust the pitch delay contour based on standard deviation and variance, allowing for dynamic adjustment of the pitch delay contour at intervals less than or equal to one block in length, and coding the adjusted contour for transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pitch delay adjustment parameters are used with fixed resolution and dynamic range, then device complexity is reduced, but long-term prediction gain deteriorates at low bit rates
Solution Approach 1:
The patent implements dynamic adjustment of pitch delay contour by introducing time-varying correction parameters that adapt to changing speech characteristics. The system updates pitch delay values across multiple sub-frames within each frame, allowing the prediction mechanism to dynamically track pitch variations rather than using a static fixed-resolution approach.
Solution Approach 2:
The patent changes the parameter representation by using differential encoding of pitch delay corrections relative to a base value. Instead of transmitting full pitch delay values with fixed resolution, the system transmits correction deltas that are scaled and biased according to the expected range of variation, effectively adapting the parameter resolution to the actual signal characteristics.
2Measurement precision
If pitch delay is coded with high resolution, then measurement precision improves, but bit rate increases
Solution Approach 1:
The patent transforms the pitch delay parameter representation by using differential encoding with biasing. The correction values are scaled to a smaller range and offset by a bias term, allowing high-precision pitch delay tracking to be achieved with fewer bits. This parameter transformation enables the system to achieve fine resolution where needed while maintaining compact representation.
Solution Approach 2:
The patent applies partial precision by using different resolution levels for different parts of the pitch delay parameter. The base pitch delay value is coded with coarse resolution, while only the necessary correction amount is coded with fine resolution. This partial application of high precision only where needed reduces the overall bit rate while maintaining measurement precision for the critical correction component.
3Manufacturing precision
If pitch delay contour is adjusted frequently, then manufacturing precision improves, but loss of energy increases
Solution Approach 1:
The patent segments the pitch delay adjustment process into hierarchical levels: frame-level base pitch delay estimation and sub-frame-level correction adjustments. This segmentation allows the system to perform coarse adjustments less frequently and fine adjustments only when necessary, reducing overall computational energy while maintaining contour accuracy through the combined effect of multiple resolution levels.
Solution Approach 2:
The patent applies adjustment actions partially by updating pitch delay contours at different rates for different portions of the signal. The base contour is updated at frame rate with lower computational cost, while correction contours are updated at sub-frame rate only for voiced segments. This partial application of frequent adjustment reduces energy consumption while maintaining accuracy where it matters most.
Data Source
AI summary
In a speech encoder/decoder a pitch delay contour endpoint modifier is employed to shift the endpoints of a pitch delay interpolation curve up or down. Particularly, the endpoints of the pitch delay interpolation curve are shifted based on a variation and/or a standard deviation in pitch delay.