Audio Encoder, Audio Decoder, Audio Encoding Method and Audio Decoding Method for Dithered Quantization for Frequency-Domain Speech and Audio Coding

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

This invention presents a hybrid audio encoding and decoding technique that improves the quality and efficiency of speech and audio signal compression, particularly at low bitrates. It does so by analyzing the spectral envelope of the signal to estimate the required bitrate for each frequency component. For spectral components needing higher bitrates, conventional arithmetic coding is used, while for components with lower bitrate needs, a dithered quantization using randomization (such as 1-bit quantization in a rotated domain) is applied. This mixed approach prevents common problems at low bitrates, such as muffled sound or excessive sparsity in the output, and provides superior perceptual quality and energy retention.

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

• Audio compression for mobile devices (e.g., smartphones, tablets) where storage and bandwidth are limited.
• Speech and audio codecs in telecommunication systems, including VoIP and mobile communications.
• Distributed recording and processing systems (e.g., wireless sensor networks, wearable technology with microphones).
• Streaming audio services seeking to maintain quality at low bandwidths.
• Hearing aids and other assistive listening devices that require efficient, high-quality audio encoding.
• Embedded audio/speech systems for IoT devices that process or transmit audio signals.
• Audio archiving where long-term storage cost and efficiency are crucial.

BenefitsContent extracted from patent full text and abstract with AI.

• Significantly improves perceived audio quality at low bitrates, reducing muffling and musical noise compared to traditional codecs.
• Retains more of the original signal's energy, especially for high-frequency components, leading to a more natural and clear sound.
• Dithered quantization and randomization reduce unwanted artifacts and prevent sparsity issues in frequency representation at low bitrates.
• Adaptively combines encoding methods for different spectral components, optimizing both quality and compression efficiency.
• The hybrid approach is computationally efficient and suited for both high-performance and resource-constrained devices (e.g., wearables, IoT).
• Provides robustness to transmission errors and is suitable for distributed or multi-device recording and processing scenarios.
• The method can be flexibly implemented in hardware or software, and is compatible with modern communication standards and codecs.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

An audio encoder for encoding an audio signal, wherein the audio signal is represented in a spectral domain, is provided. The audio encoder comprises a spectral envelope encoder (110) configured for determining a spectral envelope of the audio signal and for encoding the spectral envelope. Moreover, the audio encoder comprises a spectral sample encoder (120) configured for encoding a plurality of spectral samples of the audio signal. The spectral sample encoder (120) is configured to estimate an estimated bitrate needed for encoding for each spectral sample of one or more spectral samples of the plurality of spectral samples depending on the spectral envelope. Moreover, the spectral sample encoder (120) is configured to encode each spectral sample of the plurality of spectral samples, depending on the estimated bitrate needed for encoding for the one or more spectral samples, according to a first coding rule or according to a second coding rule being different from the first coding rule.