Optimal Quantized Linear Prediction Coefficients for Lossless Audio Compression - Scalar Quantization Revisited

Ghido F. Optimal Quantized Linear Prediction Coefficients for Lossless Audio Compression - Scalar Quantization Revisited. In: AES Convention 120. Audio Engineering Society; 2006. Paper 6757. Available from: https://aes.org/publications/elibrary-page/?id=13561

@inproceedings{Ghido2006_13561,
  author = {Ghido, Florin},
  title = {{Optimal Quantized Linear Prediction Coefficients for Lossless Audio Compression - Scalar Quantization Revisited}},
  booktitle = {AES Convention 120},
  note = {Paper 6757},
  year = {2006},
  month = may,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=13561}
}

TY  - CPAPER
TI  - Optimal Quantized Linear Prediction Coefficients for Lossless Audio Compression - Scalar Quantization Revisited
AU  - Ghido, Florin
T2  - AES Convention 120
M1  - Paper 6757
PY  - 2006
DA  - 2006/05/06
UR  - https://aes.org/publications/elibrary-page/?id=13561
PB  - Audio Engineering Society
LA  - en
AB  - Uniform scalar quantization of linear prediction coefficients with B bits is traditionally done by multiplying each coefficient with Q = 2B and rounding it to the nearest integer. We propose an improved, optimal quantization method by replacing the rounding with a more elaborated procedure. It uses on average 2 bits less per quantized prediction coefficient for a similar sum of squared errors and allows an accurate estimate of the mean squared error misadjustment as a function of Q for a given subframe and predictor order M. We introduce several efficient time-constrained probabilistic search methods for obtaining near optimal solutions. There are no required changes at the decoder and the method is applicable on a wider area of cases (mono, stereo, and multichannel prediction) than quantization of reflection coefficients. Moreover, the method enables near optimal compression for 24 bit audio using only 32 bit arithmetic operations.
ER  -