Non-Windowed DCT Audio Coding for Lower-Resource Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio compression techniques require significant resource usage due to overlapping windows in discrete cosine transform (MDCT), leading to inefficient processing and potential discontinuities at block boundaries.
Innovation Solution
Implementing back-to-back discrete cosine transforms (DCT) without windowing to transform time-domain audio signals into frequency-domain signals, reducing resource usage and discontinuities by using non-windowed DCT functions in both encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If overlapping windows in MDCT are used, then audio quality is maintained, but processing resources and time increase significantly
Solution Approach 1:
The patent extracts and removes the windowing function from the traditional MDCT process. By eliminating the windowing step and using back-to-back non-overlapping DCT transforms, the invention removes the computational overhead associated with window functions while maintaining audio quality through the modified transform approach.
Solution Approach 2:
The patent segments the audio signal into non-overlapping blocks and applies DCT transforms sequentially without overlap. This segmentation approach eliminates the need for overlapping windows while preserving audio quality through the back-to-back transform structure.
2Stability of the object's composition
If overlapping windows in MDCT are used, then discontinuities at block boundaries are suppressed, but device complexity increases
Solution Approach 1:
The patent removes the windowing function that complicates the transform process. By taking out the windowing step and using simple non-windowed DCT transforms in sequence, the invention reduces device complexity while maintaining block boundary stability through the back-to-back transform approach.
Solution Approach 2:
The patent changes the transform parameters by using non-windowed DCT transforms with specific block size configurations. This parameter modification allows the system to achieve block boundary continuity without the complexity of overlapping windows, as the back-to-back transforms naturally handle the boundary transitions.
3Productivity
If traditional quantization is used, then compression is achieved, but audio quality degrades
Solution Approach 1:
The patent modifies the quantization parameters and process to work effectively with the back-to-back DCT transform structure. By changing how quantization is applied to the transformed coefficients, the invention achieves compression while preserving audio quality that would otherwise be lost in traditional quantization processes.
Data Source
AI summary
A method including receiving a time-domain audio signal, generating a blocked time-domain audio signal as a portion of the time-domain audio signal, transforming the blocked time-domain audio signal using a first non-windowed transform function to generate a first frequency-domain audio signal, transforming the first frequency-domain audio signal using a second non-windowed transform function to generate a second frequency-domain audio signal, and compressing the second frequency-domain audio signal to generate a compressed frequency-domain audio signal.