Asymmetric Audio Window Switching for Low-Delay Frequency Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio processing technologies face challenges in achieving efficient coding of audio signals with low delay while maintaining good frequency separation, particularly in handling transitions between different transform sizes and window shapes, which often result in increased encoder look-ahead or reduced coding efficiency.
Innovation Solution
The use of asymmetric transform windows with truncated overlap portions allows for flexible transform size switching without increasing coder delay, employing a processor that derives window control signals to construct and apply windows for analysis and synthesis processing, ensuring perfect reconstruction and efficient coding of both stationary and transient signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If transition windows are inserted between frames encoded using different transform lengths, then frequency separation is improved, but encoder look-ahead is increased making it unsuitable for low-delay applications
Solution Approach 1:
The patent extracts and removes the transition window portion from the encoding process, allowing instant switching between different transform lengths without requiring transition windows. This is achieved by using a fixed overlap length for all transform sizes and selectively applying windows based on the transform length being used, thereby eliminating the time delay associated with transition windows while maintaining frequency separation through proper windowing.
Solution Approach 2:
The patent implements dynamic window selection where the window application is adapted based on the current transform length. For long transforms, windows are applied at specific positions, while for short transforms, the same window positions are used but with different overlap handling. This dynamic adaptation allows the system to switch between transform lengths instantly without requiring transition windows, resolving the contradiction between frequency separation and delay.
2Loss of time
If a fixed low window overlap is used for all transform sizes to avoid transition windows, then encoder delay is reduced, but frequency separation is degraded reducing coding efficiency for tonal signals
Solution Approach 1:
The patent applies different windowing strategies to different portions of the signal based on the transform length. For long transforms, full windows are applied to ensure good frequency separation, while for short transforms, the same window positions are used but with adjusted overlap handling. This local adaptation of window quality allows the system to maintain low delay while preserving frequency separation where needed.
Solution Approach 2:
The patent changes the window application parameters based on transform length. Instead of using a fixed low overlap for all cases, the system adjusts the effective window application by selecting different overlap portions from the same window template. This parameter change allows the system to achieve both low delay and adequate frequency separation by adapting the window usage to the specific transform length being employed.
3Loss of time
If asymmetric MDCT windows are used with shortened overlap to reduce look-ahead delay, then delay is reduced, but frequency separation is compromised
Solution Approach 1:
The patent employs asymmetric windowing where the window positions and overlap handling are different for long and short transforms. For long transforms, windows are applied in a manner that optimizes frequency separation, while for short transforms, the same asymmetric window template is used with adjusted overlap portions. This asymmetric approach allows the system to reduce delay while maintaining frequency separation by adapting the window application to the specific transform length.
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A processor for processing an audio signal (200), comprises: an analyzer (202) for deriving a window control signal (204) from the audio signal (200) indicating a change from a first asymmetric window (1400) to a second window (1402), or indicating a change from a third window (1450) to a fourth asymmetric window (1452), wherein the second window (1402) is shorter than the first window (1400), or wherein the third window (1450) is shorter than the fourth window (1452); a window constructor (206) for constructing the second window (1402) using a first overlap portion (800) of the first asymmetric window (1400), wherein the window constructor (206) is configured to determine a first overlap portion (1000) of the second window (1402) using a truncated first overlap portion of the first asymmetric window, or wherein the window constructor is configured to calculate a second overlap portion (1330) of the third window (1450) using a truncated second overlap portion (814) of the fourth asymmetric window (1452); and a windower (208) for applying the first and second windows or the third and fourth windows to obtain windowed audio signal portions (210).