Back to results
US 12,640,157 B2Granted
Higher order ambisonics encoding and decoding
AssigneeApple Inc.
InventorsMoo Young Kim, Sina Zamani, Dipanjan Sen
Life of this patent
Estimating the term…
Abstract
Encoding and decoding of higher order ambisonics, HOA, data for purposes of bitrate reduction. One aspect uses principal components analysis to produce spatial descriptors. Other aspects include various spatial descriptor quantization techniques.
Claims18
A method for encoding higher order ambisonics data, HOA data, using principal components analysis or any linear transform, the method comprising:
- subtracting a mean vector from an input HOA matrix to compute a mean subtracted HOA matrix, wherein the input HOA matrix represents a sound field of a digital audio content;
- producing a spatial descriptor, SD, by performing principal components analysis, PCA, or any linear transform based upon the mean subtracted HOA matrix;
- extracting a salient component from the mean subtracted HOA matrix;
- formatting the salient component, the SD and the mean vector into an encoded audio content bitstream;
- and
- transmitting the encoded audio content bitstream over the Internet to a decoding side of a playback system.
3 more elements
A method for decoding higher order ambisonics data, HOA data, the method comprising:
- receiving an encoded audio content bitsream over the Interet in a decoding side of a playback system;
- unformatting a salient component and a spatial descriptor, SD, from the bitstream, wherein the SD was produced by performing principal components analysis, PCA, or any linear transform based upon a mean subtracted HOA matrix;
- unformatting a mean vector from the bitsream;
- and
- computing an HOA matrix by multiplying the salient component with the SD and adding the mean vector.
2 more elements
A method for encoding higher order ambisonics data, HOA data, using principal components analysis, the method comprising:
- subtracting a mean vector from an input HOA matrix to compute a mean subtracted HOA matrix wherein the input HOA matrix represents a sound field of a digital audio content;
- producing a spatial descriptor, SD, by performing principal components analysis, PCA, or any linear transform based upon the mean subtracted HOA matrix;
- extracting a salient component directly from the input HOA matrix using the SD;
- formatting the salient component and the SD into an encoded audio content bitstream;
- and
- transmitting the encoded audio content bitstream over the Internet to a decoding side of a playback system.
3 more elements
Dependencies are inferred from the claim text.