Speech Encoder Phase Quantization for Low-Bitrate Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional speech codecs struggle in very low bitrate scenarios due to insufficient bandwidth or transmission quality issues, leading to poor performance in real-time communication.

Innovation Solution

Implement phase quantization and reconstruction techniques in speech encoding and decoding, using a linear component and weighted sum of basis functions to efficiently represent and synthesize phase values, omitting higher-frequency phase values during encoding and synthesizing them during decoding based on lower-frequency values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If conventional speech codecs are used in very low bitrate scenarios, then bandwidth consumption is reduced, but speech quality deteriorates due to insufficient bandwidth or transmission quality problems

Engineering Contradiction:
ImprovebitrateVSAvoidspeech quality
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent extracts only the essential phase information (phase values at selected frequency points) from the full spectral representation, transmitting only these critical parameters while deriving the rest at the decoder. This selective extraction maintains speech quality by preserving the most important phase relationships while minimizing transmitted data volume.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates a simplified copy of the phase spectrum by representing phase values through a compact set of parameters (phase values at selected frequency points) that can be used to reconstruct or approximate the full phase spectrum at the decoder, maintaining quality while reducing bitrate.

Inventive Principle:
Principle #26Copying

2Measurement precision

If phase values are fully transmitted to maintain accuracy, then speech quality is improved, but bandwidth consumption increases

Engineering Contradiction:
Improvephase quantization accuracyVSAvoidbandwidth
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential phase information (phase values at selected frequency points) from the full spectral representation, transmitting only these critical parameters while deriving the rest at the decoder. This selective extraction maintains speech quality by preserving the most important phase relationships while minimizing transmitted data volume.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transmits phase values at a subset of frequency points rather than all frequency points, applying partial action where the most critical phase information is transmitted while accepting approximation at other frequencies. This approach achieves sufficient precision for speech quality without the bandwidth cost of complete phase transmission.

Inventive Principle:
Principle #16Partial or excessive action

3Manufacturing precision

If all phase values are encoded and transmitted, then reconstruction accuracy is improved, but encoding complexity and transmission data increase

Engineering Contradiction:
Improvereconstruction accuracyVSAvoidencoding complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential phase information (phase values at selected frequency points) from the full spectral representation, transmitting only these critical parameters while deriving the rest at the decoder. This selective extraction maintains speech quality by preserving the most important phase relationships while minimizing transmitted data volume.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates a simplified copy of the phase spectrum by representing phase values through a compact set of parameters (phase values at selected frequency points) that can be used to reconstruct or approximate the full phase spectrum at the decoder, maintaining quality while reducing bitrate.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP4273859B1Phase quantization in a speech encoder
Publication Date: 2025.10.29 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP4273859B1 patent drawingFigure 1~2b
  • EP4273859B1 patent drawingFigure 3
  • EP4273859B1 patent drawingFigure 4

AI summary

Innovations in phase quantization during speech encoding and phase reconstruction during speech decoding are described. For example, to encode a set of phase values, a speech encoder omits higher-frequency phase values and/or represents at least some of the phase values as a weighted sum of basis functions. Or, as another example, to decode a set of phase values, a speech decoder reconstructs at least some of the phase values using a weighted sum of basis functions and/or reconstructs lower-frequency phase values then uses at least some of the lower-frequency phase values to synthesize higher-frequency phase values. In many cases, the innovations improve the performance of a speech codec in low bitrate scenarios, even when encoded data is delivered over a network that suffers from insufficient bandwidth or transmission quality problems.