Spatial Audio Cue Coding With Combined Side Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current audio encoding technologies face challenges in efficiently encoding and synthesizing complex auditory scenes with multiple audio sources, as they often require significant computational resources and bitrate, and struggle to maintain spatial cues accurately across different playback formats.
Innovation Solution
The method involves generating combined cue codes from multiple audio channels, which are then downmixed and encoded into fewer transmitted channels, allowing for efficient transmission and decoding of spatial cues like inter-channel time difference (ICTD), inter-channel level difference (ICLD), and inter-channel correlation (ICC) data, enabling accurate synthesis of multi-channel audio signals with reduced bitrate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple audio channels are encoded with full spatial cues, then spatial audio quality is improved, but bitrate and computational resources increase significantly
Solution Approach 1:
The patent extracts only the essential spatial cue information (ICTD, ICLD, ICC) from multiple audio channels and transmits this compact side information separately from the downmixed audio signal. This allows the full spatial cues to be reconstructed at the decoder without transmitting all original channel data, thereby reducing bitrate while maintaining spatial accuracy.
Solution Approach 2:
The patent transforms the representation of spatial audio by changing from transmitting multiple full audio channels to transmitting a downmixed signal accompanied by compact parametric side information. This parameter-based representation (using ICTD, ICLD, ICC values) significantly reduces the data quantity while preserving the essential spatial characteristics needed for accurate cue reconstruction.
2Measurement precision
If multiple audio channels are encoded with full spatial cues, then spatial audio quality is improved, but computational resources increase significantly
Solution Approach 1:
The patent extracts only the essential spatial cue information (ICTD, ICLD, ICC) from multiple audio channels and transmits this compact side information separately from the downmixed audio signal. This allows the full spatial cues to be reconstructed at the decoder without transmitting all original channel data, thereby reducing bitrate while maintaining spatial accuracy.
Solution Approach 2:
The decoder uses the transmitted compact side information (ICTD, ICLD, ICC parameters) to self-generate the spatial cues needed for reconstructing the original multi-channel audio signal. This eliminates the need for the decoder to store or process multiple full-channel signals, reducing computational complexity while maintaining spatial accuracy.
3Quantity of substance
If audio is downmixed to fewer channels for transmission, then bitrate is reduced, but spatial information is lost
Solution Approach 1:
The encoder performs preliminary extraction and transmission of spatial cue parameters (ICTD, ICLD, ICC) before the actual audio transmission. By preparing and transmitting this side information in advance alongside the downmixed signal, the decoder can reconstruct the spatial cues without losing essential spatial information, even though the audio itself is downmixed to fewer channels.
Solution Approach 2:
The patent introduces compact parametric side information (ICTD, ICLD, ICC values) as an intermediary that carries spatial cue data between the encoder and decoder. This intermediary representation preserves the essential spatial information that would otherwise be lost in downmixing, enabling accurate spatial reconstruction at the decoder without requiring full multi-channel transmission.
Data Source
AI summary
At an audio encoder, cue codes are generated for one or more audio channels, wherein a combined cue code (e.g., a combined inter-channel correlation (ICC) code) is generated by combining two or more estimated cue codes, each estimated cue code estimated from a group of two or more channels. At an audio decoder, E transmitted audio channel(s) are decoded to generate C playback audio channels. Received cue codes include a combined cue code (e.g., a combined ICC code). One or more transmitted channel(s) are upmixed to generate one or more upmixed channels. One or more playback channels are synthesized by applying the cue codes to the one or more upmixed channels, wherein two or more derived cue codes are derived from the combined cue code, and each derived cue code is applied to generate two or more synthesized channels.