Time-Warped Audio Transform Coding for Smooth Frame Reconstruction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio coding techniques face inefficiencies in encoding signals with pitch variations, leading to increased bit rates and audible discontinuities due to uncontrollable duration and bandwidth limitations in time-warped signals, as well as significant bit-rate overhead for transmitting pitch information.

Innovation Solution

A method for efficient audio coding using time warping, where a common time warp is estimated for consecutive frames to facilitate block transforms, allowing for overlap and add reconstruction without introducing rapid pitch variations at frame borders, and reducing the need for additional pitch information transmission by using warp parameters instead of pitch parameters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If simple time warping is applied to achieve constant pitch over voiced segments, then coding efficiency is improved, but uncontrollable duration and bandwidth limitations occur leading to violation of signal duration and bandwidth constraints

Engineering Contradiction:
Improvecoding efficiencyVSAvoidsignal duration
Core Design Contradiction:
ProductivityVSDuration of action of moving object

Solution Approach 1:

The patent changes the parameter representation from absolute pitch values to relative pitch ratios between adjacent frames. This allows the time-warping operation to maintain constant pitch within frames while the sequence of pitch ratios controls the overall pitch contour, thereby maintaining signal duration and bandwidth constraints while achieving coding efficiency.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent segments the signal processing into frame-level constant pitch transformation and inter-frame pitch ratio transmission. Each frame is processed independently with time-warping to achieve local stationarity, while the global duration and bandwidth constraints are maintained through controlled pitch ratio sequences between frames.

Inventive Principle:
Principle #1Segmentation

2Duration of action of moving object

If additional side information on tape speed (pitch) is transmitted to control duration, then signal duration is preserved, but substantial bit-rate overhead is introduced especially at low bit-rates

Engineering Contradiction:
Improvesignal durationVSAvoidbit-rate
Core Design Contradiction:
Duration of action of moving objectVSQuantity of substance

Solution Approach 1:

The patent transmits pitch ratios between adjacent frames instead of absolute pitch values or tape speed information. This parameter transformation reduces the bit-rate overhead because pitch ratios can be represented with fewer bits and exhibit better predictive coding properties, while still allowing accurate reconstruction of the signal duration.

Inventive Principle:
Principle #35Parameter changes

3Duration of action of moving object

If consecutive non-overlapping segments are processed independently by time warp to preserve duration, then signal duration is maintained, but pitch jumps at segment boundaries introduce audible discontinuities and loss of coding efficiency

Engineering Contradiction:
Improvesignal durationVSAvoidaudible discontinuities
Core Design Contradiction:
Duration of action of moving objectVSObject-affected harmful factors

Solution Approach 1:

The patent applies time-warping transformation preliminarily within each frame to achieve local pitch stationarity before encoding. By processing overlapping or adjacent frames with continuous pitch ratio control, the harmful pitch jumps at boundaries are eliminated while maintaining signal duration through controlled pitch evolution between frames.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP1953738B1Time warped modified transform coding of audio signals
Publication Date: 2016.10.12 DOLBY SWEDEN AB
  • EP1953738B1 patent drawingFigure 1
  • EP1953738B1 patent drawingFigure 2~2b
  • EP1953738B1 patent drawingFigure 3a~3b

AI summary

A spectral representation of an audio signal having consecutive audio frames can be derived more efficiently, when a common time warp is estimated for any two neighbouring frames, such that a following block transform can additionally use the warp information. Thus, window functions required for successful application of an overlap and add procedure during reconstruction can be derived and applied, the window functions already anticipating the re-sampling of the signal due to the time warping. Therefore, the increased efficiency of block-based transform coding of time-warped signals can be used without introducing audible discontinuities.