Compressed HOA Sound Decoding With Minimum-Bit Gain Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high bit rate required for transmitting Higher Order Ambisonics (HOA) representations poses a challenge for practical applications, particularly in streaming, due to the quadratic growth of expansion coefficients with order N, and existing compression methods lack efficient coding of non-differential gain values without specifying constraints on value ranges.
Innovation Solution
An inter-relation is established between the value range of input HOA representations and potential maximum gains to determine the minimum bits required for encoding non-differential gain values, using a normalised HOA data frame representation and specific matrix operations to derive the lowest integer number of bits needed for compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If HOA representation is transmitted with high quality (order N=4, 16 bits per sample), then sound quality is improved, but bit rate increases to 19.2 MBits/s which is too high for practical streaming applications
Solution Approach 1:
The HOA representation is segmented into a directional component and a residual ambient component. The directional component contains prominent sound sources and is coded with higher precision, while the ambient component contains diffuse background sound and is coded with lower precision. This segmentation allows the total bit rate to be reduced while maintaining perceptual quality, as the human ear is more sensitive to directional sound sources than to ambient background noise.
Solution Approach 2:
Different parts of the HOA representation are assigned different quality levels. The directional component, which contains important spatial information about prominent sound sources, is encoded with higher bit allocation and better quality. The ambient component, which contains less critical background information, is encoded with lower bit allocation and acceptable quality. This local differentiation of quality allows overall bit rate reduction while preserving the most important auditory information.
2Productivity
If compression is applied to reduce bit rate, then transmission efficiency is improved, but additional side information and gain control processing are required which increases system complexity
Solution Approach 1:
Gain control processing is applied preliminarily to the HOA coefficients before compression encoding. The gain values are calculated in advance to normalize the amplitude of the directional and ambient components, ensuring that the compressed representation maintains correct relative amplitude relationships during decompression. This preliminary gain control simplifies the compression process by preventing overflow and underflow issues, reducing the need for complex protection mechanisms during transmission.
Solution Approach 2:
The system uses differential gain values that represent the ratio between ambient and directional component amplitudes. These gain values are encoded as side information and fed back to the decoder to reconstruct the original amplitude relationships. The feedback mechanism ensures that the decompressed HOA representation accurately reproduces the spatial and amplitude characteristics of the original sound field, maintaining fidelity while using compression.
3Quantity of substance
If differential gain values are used for compression, then bit usage is optimized, but random access becomes problematic due to temporal dependencies between frames
Solution Approach 1:
The bitstream is segmented into independent access units, each containing a complete set of HOA coefficients and associated gain values for a specific time frame. This segmentation breaks the temporal dependency chain, allowing any access unit to be decoded independently without requiring previous frames. The differential gain values are calculated and stored within each access unit relative to a reference frame, enabling random access while maintaining compression efficiency.
Solution Approach 2:
Reference frames are selected in advance at regular intervals throughout the bitstream, and differential gain values are calculated relative to these predetermined references. This preliminary establishment of reference points allows the decoder to jump to any access unit and reconstruct the sound field by using the stored differential gains without needing to process preceding frames, thus enabling random access while preserving the bit-saving benefits of differential encoding.
Data Source
AI summary
When compressing an HOA data frame representation, a gain control (15, 151) is applied for each channel signal before it is perceptually encoded (16). The gain values are transferred in a differential manner as side information. However, for starting decoding of such streamed compressed HOA data frame representation absolute gain values are required, which should be coded with a minimum number of bits. For determining such lowest integer number (βe) of bits the HOA data frame representation (C(k)) is rendered in spatial domain to virtual loudspeaker signals lying on a unit sphere, followed by normalisation of the HOA data frame representation (C(k)). Then the lowest integer number of bits is set toβe=⌈log2(⌈log2(KMAX·O)⌉+1)⌉.