Audio Signal Encoding with Amplitude Sorting for Low-Bitrate Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional audio coding methods face challenges in maintaining signal quality at low bit rates, particularly with lossy coding methods that introduce perceptible artifacts and require significant computing power, while lossless methods struggle with data reduction and efficient encoding.
Innovation Solution
A device and method for encoding and decoding audio signals that involves sorting the input signal by amplitude, followed by prediction and entropy coding, allowing for effective coding of sorted data and re-sorting rules, which can include permutation and inversion tables, to achieve efficient data representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If lossy coding methods with psychoacoustic models are used, then data reduction is achieved, but signal quality deteriorates with perceptible coding artifacts at low bit rates
Solution Approach 1:
The patent applies preliminary sorting of spectral coefficients by amplitude before coding. This preliminary action reorganizes the data so that energy is concentrated in fewer coefficients, enabling more efficient encoding. By sorting beforehand, the system can achieve better data reduction while maintaining signal quality, as the most significant coefficients are encoded with higher precision while less significant ones can be coarsely quantized or discarded.
2Loss of information
If conventional transformation-based coding is used, then frequency domain representation is achieved, but computing power consumption increases due to quantization and downsampling operations
Solution Approach 1:
The patent extracts and encodes only the most significant spectral coefficients after sorting, rather than processing all coefficients equally. By taking out only the essential information (top N sorted coefficients), the system reduces computing power requirements for quantization and encoding operations while maintaining the frequency domain representation quality. This selective extraction avoids unnecessary processing of less significant coefficients.
3Manufacturing precision
If lossless coding methods are used, then perfect reconstruction is achieved, but data reduction efficiency decreases compared to lossy methods
Solution Approach 1:
The patent applies different coding strategies to different portions of the sorted spectral coefficients. The most significant coefficients (at the beginning of the sorted sequence) are encoded with lossless precision to ensure perfect reconstruction of important signal components. Less significant coefficients (toward the end of the sequence) can be encoded with lossy compression or discarded entirely. This local differentiation of quality levels enables both perfect reconstruction of critical information and overall data reduction.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
The invention relates to a device (100) for encoding a sequence of scanning values of an audio signal, wherein each scanning value within the sequence has an original position. The device (100) comprises a system (110) for sorting the scanning values according to size in order to obtain a sorted sequence of scanning values, wherein each scanning value has a sorting position within the sorted sequence. Furthermore, the device (100) comprises a system (120) for encoding the sorted scanning values and information about a relationship between the original and sorting positions of the scanning values.