Lossless Predictive Audio Compression

Lossless Predictive Audio Compression ( LPAC; German lossless predictive audio data compression) is a lossless compression method for audio. Use it finds in a lossless audio format with the file name extension. Pac. It was developed by Tilman Liebchen from the Communication Systems of the Technical University of Berlin.

Meanwhile, there is a historical format because it is no longer being developed and in MPEG -4 Audio Lossless Coding has his official successor.

Existing features

The method provides in comparing mediocre pack rates and decodes very quickly. Supports mono and stereo PCM data in 8 -, 12 -, 16 -, 20 - and 24-bit resolution samples. The compression distinguishes between six compression modes which differ by compression strength and speed. Furthermore, it was in LPAC a Cyclic Redundancy Check ( engl. cyclic redundancy check) implemented.

Features missing

It lacks some modern features:

  • Tagging ( embedding of additional information about the title, author, etc.)
  • Streaming
  • Multi-channel audio
  • Replay Gain standards
  • Error correction mechanisms ( CRC checksums make defects at least recognizable)

Additional Features

  • Rapid light to any position in a file, but which must first be activated by an additional option, which might then result in larger file
  • Proprietary format

Practical additional options such as optional separate storage of prediction signal and an error correction file or self-extracting files (see WavPack ) are also absent.

License, source code

Graphically -use as well as command -line utilities for Windows, Linux and Solaris, as well as a program library (DLL) are available for free. However, the source code is not freely available. Born out LPAC MPEG -4 Audio Lossless Coding is - as an open standard - disclosed and described.

History

LPAC is the successor of Lossless Transform Audio Compression ( LTAC ). Late 2004/early 2005, the development of LPAC was set because it was in a revised version in December 2005 for MPEG -4 Audio Lossless Coding ( ALS) standard. It also important previously missing features were added and the format has been disclosed.

Technology

Using adaptive linear prediction, the audio signal is predicted. The residual signal, comprising the deviations of the prediction, then passes through a Rice entropy coding, takes advantage of the different probabilities of occurrence of samples of the residual signal.

Proprietary software Freeware Audioformat
529417
de