AU Full Form
AU is an audio brief, a standard audio format for UNIX sound files. The AU file format is the standard Java programming language audio file format which was first created by Sun Microsystems. Usually, AU files end with a .au extension. It was originally headerless since it was simply 8-bit μ-law-encoded data with a sample rate of 8000 Hz. Other sample rates are often used by hardware suppliers up to 8192 Hz, often integer several signal frequencies for video clocks. More recent files have six unsigned 32-bit words in their headers, an optional chunk of information, and then a chunk of data.
While the format now supports many formats of audio encoding, the logarithmic encoding remains linked to μ-law. This encoding was originated from hardware SPARCstation 1, where SunOS exposed the application encoding to applications via the /dev/audio interface. The encoding and interface have become the Unix sound standard.
AU file format:
- File Description: AU or NeXT/Sun sound file
- File Extension: Commonly .au or .snd
- File Byte Order: Big-endian
- Type of format: audio file format, container format
- Container for: Audio, most often μ-law
The encoding type is dependent on the value of the field “encoding” (word 3 of the header). Formats 2 to 7 are linear and therefore technically lossless PCM uncompressed (although not necessarily free of quantization error, especially in an 8-bit form). Formats 1 and 27 respectively are μ-law and A-law, both logarithmic PCM companding representation and probably loss as they pack what would otherwise be nearly sixteen dynamic bits into eight bits of encoded data, although this is achieved with an altered dynamic response and there is no data that is actually “thrown away”.
Formats 23 to 26 are ADPCM, an early form of loss compression, usually with 4 bits of encoded data per audio quantitative (for 4:1 efficiency with 16-bit input, or 2:1 with 8-bit; equivalent to e.g. MP3 encoding of CD-quality with a low-quality encoder rate of 352kbit). Many other commands are DSP and/or data designed to be processed using the software NeXT Music Kit.
A variable-length field is the header structure. The contents of this field are currently undefined unless its length is several bytes long and at least one null byte must be completed. The audio data segment starts right after the annotation field at an 8-byte limit. Audio data are encoded in the file header format. The current version only supports one audio data segment per file. Most audio applications are currently unaware of the variable length annotation field.
Note: The signed data is encoded in PCM formats
Samples of the 16-bit stereo data files are as follows. The following are examples:
- M-Al-A.au (46 kB) –
AU file, stereo A-law data.
- M1-m-P.au (46 kB) –
AU file, stereo µ-law data.