NAME

adpcm_enc, adpcm_dec − encode or decode voice-quality audio files

SYNOPSIS

adpcm_enc [ filename ... ]

adpcm_dec [ filename ... ]

DESCRIPTION

adpcm_enc compresses audio data using the CCITT G.721 ADPCM voice data encoding scheme.  This data compression technique reduces the storage requirements of voice-quality audio data by 50%.  adpcm_dec decodes ADPCM audio files to standard u−law audio format. 

adpcm_enc reads samples from the named audio files and encodes them into 4-bit ADPCM codes.  The 4-bit codes are packed in bytes, lower nibble first, and written to standard output, preceded by an audio file header.  If no input files are specified, the standard input stream is read.  If any of the input files are ADPCM encoded audio files already, they are copied without change. 

adpcm_dec reverses the processing of adpcm_enc by decoding ADPCM samples from the named audio files.  The output is written to the standard output as a u−law audio data file, doubling the data storage requirements.  If no input files are specified, the standard input stream is read.  If any of the input files are u−law coded audio files already, they are copied without change. 

For audio playback, ADPCM encoded files must first be decoded.  For example,

example% adpcm_dec file | /usr/demo/SOUND/play

ERRORS

Input files must conform to the audio file format, otherwise an error message is written to standard error and no output is generated. 

SEE ALSO

audiotool(1), audio(4), raw2audio(6), record(6), play(6). 

CAVEATS

The current implementation of adpcm_enc encodes u−law audio files only.  Future revisions may support other audio encoding schemes, e.g., linear PCM, A-law, G.722, G.723, etc. 

The CCITT G.721 ADPCM algorithm is particularly oriented towards encoding voice data.  If these programs are used to compress music or other types of sounds, there may be some degradation in the resulting signal quality. 

Solbourne Computer, Inc.  —  3/7/91