Audio Decoding Apparatus Reducing Multiplication Processes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for downmixing multi-channel audio signals to stereo signals are inefficient due to the high number of multiplication processes required, which can slow down processing speed, especially when a CPU performs additional tasks simultaneously.
Innovation Solution
A decoding and encoding apparatus that transforms encoded audio signals into time-domain transform block-based signals, multiplies them by a product of mixture ratios and window functions, and synthesizes these signals to reduce the number of multiplication processes needed during downmixing, allowing for faster processing without increasing the complexity of calculations when changing window functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the conventional downmixing process is used where decoded multi-channel audio signals are multiplied by mixture ratio coefficients and summed, then the downmixing can be performed with separate multiplication and mixing steps, but the number of multiplication processes increases significantly, slowing down processing speed
Solution Approach 1:
The patent combines the window function multiplication and mixture ratio multiplication into a single operation by pre-calculating a composite window function that incorporates both the original window function and the mixture ratio coefficients. This merging eliminates separate multiplication steps and reduces computational complexity while maintaining the same downmixing functionality.
Solution Approach 2:
The patent performs preliminary calculation of the composite window function by combining the window function and mixture ratio coefficients before the actual downmixing process. This pre-processing step stores the combined parameters, so that during runtime, only simple multiplication and addition operations are needed without recalculating the complex coefficient combinations.
2Productivity
If the CPU performs downmixing process simultaneously with other tasks, then system resource utilization is improved, but the processing speed of audio signals is lowered due to the high number of multiplication processes
Solution Approach 1:
The patent changes the parameters by pre-calculating and storing composite window functions that incorporate mixture ratio coefficients. This parameter transformation converts complex runtime multiplication operations into simpler operations using pre-computed values, reducing processing time and enabling parallel execution with other CPU tasks.
3Manufacturing precision
If window functions are changed during processing to adapt to different audio characteristics, then audio quality is improved, but the amount of calculation increases when using the conventional separate multiplication approach
Solution Approach 1:
The patent creates a universal composite window function that serves multiple purposes: it applies the original window function for signal processing and simultaneously incorporates the mixture ratio coefficients for downmixing. This multi-functional approach allows window function changes without increasing calculation amount, as the composite structure handles both functions in a single operation.
Data Source
AI summary
A decoding apparatus (10) is disclosed which includes: a storing means (11) for storing encoded audio signals including multi-channel audio signals; a transforming means (40) for transforming the encoded audio signals to generate transform block-based audio signals in a time domain; a window processing means (41) for multiplying the transform block-based audio signals by a product of a mixture ratio of the audio signals and a first window function, the product being a second window function; a synthesizing means (43) for overlapping the multiplied transform block-based audio signals to synthesize audio signals of respective channels; and a mixing means (14) for mixing audio signals of the respective channels between the channels to generate a downmixed audio signal. Furthermore, an encoding apparatus is also disclosed which downmixes the multi-channel audio signals, encodes the downmixed audio signals, and generates the encoded, downmixed audio signals.