Sound Frame Length Adaptation for Transform Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound synthesis methods in the transform domain are inefficient due to frame lengths not being powers of two, leading to redundant data and increased processing time, especially when converting between time and frequency domains.
Innovation Solution
A device and method that form frames with sound parameters, transform them into efficient inverse frames, and either discard or repeat frames to match the desired time domain length, utilizing nearly all data for improved efficiency and maintaining sound duration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If frame length is chosen to be a power of two for efficient transform, then transform efficiency is improved, but time domain sound duration becomes inaccurate
Solution Approach 1:
The patent divides the sound signal into multiple frames, where each frame is processed independently through the transform domain. This segmentation allows the use of power-of-two frame lengths for efficient FFT/IFFT transforms while maintaining the ability to reconstruct accurate time domain sound duration through proper frame management and overlap-add operations.
Solution Approach 2:
The patent dynamically adjusts frame parameters including frame length, overlap factor, and shift amount to optimize both transform efficiency and time domain accuracy. By changing these parameters adaptively, the system achieves efficient power-of-two transforms while maintaining precise time domain representation of the sound signal.
2Loss of time
If frame length is increased to improve transform efficiency, then processing time is reduced, but redundant data increases
Solution Approach 1:
The patent uses a window function that extends slightly beyond the actual required frame length, creating a small amount of redundant data at the edges. This excessive action allows the use of efficient power-of-two transform lengths while the windowing function ensures that the redundant portions contribute minimally to the final reconstructed signal, thus reducing overall processing time without significant loss of information.
Solution Approach 2:
The patent extracts and removes the redundant portions of the transformed data through careful windowing and overlap-add operations. By taking out the unnecessary edge portions that result from padding to power-of-two lengths, the system maintains transform efficiency while minimizing the amount of redundant data that needs to be processed and stored.
3Productivity
If all inversely transformed data is used to maximize efficiency, then transform efficiency is improved, but frame length mismatch occurs
Solution Approach 1:
The patent implements dynamic frame length adjustment where the actual output frame length is determined by the transform length minus the window function support. This dynamic approach allows the system to adapt to different transform lengths (including power-of-two lengths for efficiency) while automatically adjusting the effective frame length to maintain compatibility with the desired time domain duration through overlap-add processing.
Data Source
Figure 1~2
Figure 3~4
Figure 5~7
AI summary
A method of producing time domain sound data (B) from sound parameters (A), the method comprising the steps of: forming first frames, each first frame containing sound parameters representing sound, - forming second frames from the first frames, each second frame containing transform domain sound data derived from the sound parameters, the transform domain sound data of each second frame representing sound having a specific time domain length, and each second frame having a length corresponding with an efficient inverse transform, inversely transforming the second frames into third frames (G1, G2,...), each third frame containing time domain sound data corresponding to the transform domain sound data of a second frame, and each third frame having a length equal to a second frame, outputting substantially all time domain sound data (B) of each third frame, and discarding or repeating first frames (F3, F7) as necessary to compensate for any length difference between the said specific time domain length (P) and the length of the third frames (Q).