Audio Signal Encoding Using Phase Envelope Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for encoding sinusoidal model parameters, particularly phase parameters, are inefficient and require high computational complexity, making them unsuitable for implementation using fixed-point arithmetic.
Innovation Solution
The method involves obtaining an initial discrete spectral representation of an audio frame, estimating a phase envelope, calculating a residual representation, and encoding it using a linear combination of codewords, allowing for efficient encoding of sinusoidal model parameters using fewer bits and suitable for both floating-point and fixed-point arithmetic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques are used for encoding sinusoidal model parameters, then the encoding can be performed, but the computational complexity is high and it requires more bits
Solution Approach 1:
The patent segments the spectral representation into multiple discrete spectral representations (DSRs), each covering a different frequency range. By dividing the encoding task into multiple segments that can be processed independently and in parallel, the computational complexity is reduced while maintaining encoding precision.
Solution Approach 2:
The patent replaces complex mechanical/computational processing with a mathematical approach using discrete spectral representations and frequency-domain operations. This substitution of the processing mechanism reduces computational complexity while preserving the ability to accurately encode sinusoidal model parameters.
2Measurement precision
If conventional techniques are used for encoding sinusoidal model parameters, then the encoding can be performed, but the bit budget is high
Solution Approach 1:
By segmenting the spectral representation into multiple DSRs covering different frequency ranges, the patent enables more efficient bit allocation. Each segment can be encoded with appropriate precision, reducing the total bit budget while maintaining overall parameter encoding accuracy.
Solution Approach 2:
The patent transforms the encoding approach by changing from direct time-domain parameter encoding to frequency-domain discrete spectral representation. This parameter transformation allows for more compact representation of sinusoidal model parameters, reducing the bit budget required while preserving encoding accuracy.
3Ease of operation
If conventional techniques are used for encoding phase parameters, then the encoding can be performed, but it is inefficient and unsuitable for fixed-point arithmetic
Solution Approach 1:
The patent replaces complex phase parameter encoding operations with discrete spectral representation operations that are naturally suited for fixed-point arithmetic. By working in the frequency domain with DSRs, the system achieves encoding efficiency while being compatible with fixed-point implementations.
Solution Approach 2:
The patent changes the representation parameters from direct phase parameters to discrete spectral representations. This parameter transformation simplifies the arithmetic operations required, making the encoding process more efficient and suitable for fixed-point arithmetic implementations.
Data Source
AI summary
Some embodiments relate to techniques for encoding an audio signal represented by a plurality of frames including a first frame. The techniques include using at least one computer hardware processor to perform: obtaining an initial discrete spectral representation of the first frame; obtaining a primary discrete spectral representation of the initial discrete spectral representation at least in part by estimating a phase envelope of the initial discrete spectral representation and evaluating the estimated phase envelope at a discrete set of frequencies; calculating a residual discrete spectral representation of the initial discrete spectral representation based on the initial discrete spectral representation and the primary discrete spectral representation; and encoding the residual discrete spectral representation using a plurality of codewords.