Audio Time-Scale Modification Using Dynamic Segment Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing time-scale modification algorithms, such as PICOLA, assume periodic waveforms and may fail to accurately determine starting points, leading to inaccuracies in audio signal compression and expansion due to non-periodic nature of real-world audio signals.
Innovation Solution
A system and method that select a proper time length and starting point to minimize the difference between adjacent segments, using triangle window functions for overlap-add operations to generate new segments for compression or expansion, allowing for flexible selection of time lengths and starting points to ensure the difference is below a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing time-scale modification algorithms (such as PICOLA) are used, then the audio signal can be compressed or expanded, but the accuracy deteriorates due to the periodic waveform assumption failing for non-periodic real-world audio signals
Solution Approach 1:
The patent changes the fundamental parameter assumption from periodic to non-periodic waveforms. Instead of assuming fixed periodic intervals between segments, the algorithm dynamically determines segment boundaries and time lengths based on actual waveform characteristics, allowing accurate time-scale modification for both periodic and non-periodic audio signals
Solution Approach 2:
The patent introduces dynamic adaptation by allowing the starting point and time length of segments to vary based on the specific audio signal being processed. The algorithm dynamically selects optimal segment parameters rather than relying on fixed periodic assumptions, making the system adaptable to different waveform types including non-periodic signals
2Device complexity
If fixed periodic segment selection is used, then the processing is simple, but the measurement precision deteriorates in determining starting points for non-periodic signals
Solution Approach 1:
The patent performs preliminary analysis of the waveform to identify optimal starting points and segment boundaries before executing the time-scale modification. By pre-processing the signal to detect characteristic features and determine appropriate segment parameters, the algorithm achieves high precision in starting point determination without excessive computational complexity during the main processing stage
Data Source
AI summary
System and methods are provided for modifying audio signals. A waveform representing an audio signal changing over time is received. A first time length is selected. A first starting point in the waveform is selected. A first pair of adjacent segments of the waveform are determined based at least in part on the first starting point and the first time length. The first pair of adjacent segments each correspond to the first time length. A first difference measure associated with the first pair of adjacent segments is calculated. In response to the first difference measure being smaller than a threshold, compression or expansion of the waveform is performed based at least in part on the first time length and the first starting point.