Apparatus and Method for Processing an Audio Signal Using a Combination in an Overlap Range

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention relates to an apparatus and method for processing audio signals, particularly for reducing or eliminating artifacts known as time-domain aliasing that can appear when modifying audio in the frequency domain. The technique involves processing overlapping blocks of spectral (frequency domain) data from an audio signal with different modification values (such as varying gain), then intelligently combining the resulting signals within the overlap region, usually using a cross-fade or error correction, to produce a clean, aliasing-free audio output.

Use CasesContent extracted from patent full text and abstract with AI.

• Audio compression codecs (such as MP3, AAC, etc.) to improve sound quality after spectral processing or gain adjustments
• Post-processing in audio editing software to apply effects or enhancements without unwanted artifacts
• Speech codecs for communication devices, ensuring clear voice reproduction after dynamic adjustments
• Music production and mastering, particularly during dynamic range compression, equalization, or bandwidth extension
• Hearing aids or assistive listening devices to ensure modifications do not introduce aliasing noise
• Any system performing spectral band replication (BWE) or bandwidth filling in digital audio processing

BenefitsContent extracted from patent full text and abstract with AI.

• Reduces or eliminates unwanted time-domain aliasing artifacts after frequency-domain processing, improving audio clarity
• Allows for more flexible and sophisticated spectral modifications (like gain changes, dynamic range control, or bandwidth extension) without introducing audible errors
• Maintains high audio quality even when time- or frequency-varying modifications are applied
• Enables efficient implementation with minimal computational complexity and delay compared to prior art solutions
• Provides a general framework that can be integrated into various digital audio codecs and processing pipelines
• Supports advanced bandwidth extension/filling techniques for efficient audio compression and improved perceived quality

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

An apparatus for processing an audio signal comprising a sequence of blocks (114) of spectral values comprises: a processor (100) for processing the sequence of blocks using at least one modification values (102) for a first block to obtain aliasing-reduced or aliasing-free first result signal in an overlap range (170) and using at least one second different modification value (106) for a second block of the sequence of blocks to obtain an aliasing-reduced or aliasing-free second result signal (108) in the overlap range (170); and a combiner (110) for combining the first result signal (104) and the second result signal (108) in the overlap range (170) to obtain a processed signal (112) for the overlap range (170).