Sustainability of Digital Formats: Planning for Library of Congress Collections

Introduction | Sustainability Factors | Content Categories | Format Descriptions | Contact
Format Description Categories >> Browse Alphabetical List

AES3, Digital Audio Interface Format

>> Back
Table of Contents
Format Description Properties Explanation of format description terms

Identification and description Explanation of format description terms

Full name Formal name: AES3-2003. AES standard for digital audio — Digital input-output interfacing — Serial transmission format for two-channel linearly represented digital audio data. Common name: AES3.

Audio Engineering Society interface format for serial digital transmission of stereo or two-channel LPCM (Linear Pulse Code Modulated) sound. The data is sent in audio blocks, each of which is made up of 192 frames numbered 0 to 191. Each frame is divided in 2 subframes (or channels): A (left) and B (right). Each subframe contains the information for one single sample of the PCM audio. LCPM data is "expected" to be sampled at the preferred frequencies recommended in AES5; see Notes below.

A method for inserting standardized unique identifiers into the AES3 stream is specified in AES52-2006, specifically covering the UMID (Unique Material Identifier specified in SMPTE 330M-2000) and the UUID (Universally Unique Identifier as documented in ISO/IEC 11578:1996 and IETF RFC 4122).

An extension of the AES3 interface has been developed by the Society of Motion Picture and Television Engineers (AES3_SMPTE), and it specifies a method for using the interface to transmit non-PCM AC3 compressed audio and other data.

The main thrust of AES3 concerns cabling and the physical interface between devices (see Notes below); this Web site is primarily devoted to formats pertaining to the storage of digital content bitstreams and to related file, wrapper, or bundling formats.

Note: Revised versions of AES3 and AES55 published in 2009 and 2007 permit the use of MPEG surround sound; more information will be added to this Web site in the future.

Production phase Interface format that may be used at any point in a production life cycle.
Relationship to other formats
    Contains LPCM, Linear Pulse Code Modulated Audio
    Has earlier version AES3-1985 and -1992 (also published as ANSI S4.40-1985 and -1992), not documented here.

Local use Explanation of format description terms

LC experience or existing holdings See LPCM
LC preference See LPCM

Sustainability factors Explanation of format description terms

Disclosure Open standard. Developed by the Audio Engineering Society.
    Documentation AES3-2003. AES standard for digital audio — Digital input-output interfacing — Serial transmission format for two-channel linearly represented digital audio data.
Adoption Widely adopted.
    Licensing and patents None.
Transparency See LPCM
Self-documentation Not applicable. Metadata can be embedded in some file formats that incorporate LPCM bitstreams.
External dependencies The specification covers wiring and cable connectors.
Technical protection considerations None

Quality and functionality factors Explanation of format description terms

Normal rendering Good support.
Fidelity (high audio resolution) Fidelity considerations reflect the use of linear PCM encoding; see LPCM.
Multiple channels AES3 (2003) supports the interleaving of two LPCM streams.
Support for user-defined sounds, samples, and patches Not applicable
Functionality beyond normal rendering Not applicable

File type signifiers and format identifiers Explanation of format description terms

Tag Value Note
Filename extension See note.  Generally depends upon the wrapper used for the "storage" version of LPCM.  LPCM data may be stored in "raw" form, with ".raw" as a possible file extension.  A variant of this encoding is captured by devices manufactured by Snell and Wilcox, reportedly carrying an ".aes" extension.
Internet Media Type Not applicable.   
Magic numbers Not found.  Comments welcome.   

Notes Explanation of format description terms


The main body of AES3 specifies signal transmission via a single shielded twisted wire pair. Annex D, however, discusses the carriage of AES3 signals on structured wiring, e.g., the "category 5" cables associated with computer network installations, a topic that receives extended treatment in AES47, which specifies the method of carrying multiple channels of audio in linear PCM or AES3 format across an asynchronous transfer mode (ATM) network. Elaborations on audio signal transfer are the topics of AES10-2003, AES Recommended Practice for Digital Audio Engineering 8221; Serial Multichannel Audio Digital Interface (MADI) (Revision of AES10-1991), and AES50-2005: AES standard for digital audio engineering 8221; High-resolution multi-channel audio interconnection.

Certain sampling frequencies are recommended by the AES in AES5, pp. 5-6:

  • The recommended sampling frequency for digital audio encoding shall be 48 kHz ± 10 parts per million. This frequency is compatible with television and motion picture systems, and it permits the encoding of audio programs with full 20-kHz bandwidth. It is recommended for all applications of program origination, processing, and interchange. To maintain synchronization whenever digital audio using the standard 48-kHz sampling frequency is utilized in conjunction with television or motion pictures, the average number of samples per frame shall be [as follows: nominal frame rate of 24, then 2,000 samples per frame; 25 frames, then 1,920 samples; 30 frames, 1,600 samples; and 30/1,001 frames, 8008/5 samples].
  • For an application directly related to certain consumer products, a sampling frequency of 44,1 kHz may be used. This application may include the interchange of program material prepared specifically for such products.
  • For broadcast and transmission-related applications with restricted channel capacity and where a nominal audio bandwidth of 15 kHz is considered adequate, a sampling frequency of 32 kHz may be used, in accordance with ITU-T Recommendation J.53.
  • For applications with an audio bandwidth greater than 20 kHz or in order to permit the use of a wider transition region in the anti-alias filtering a rate of 96 kHz ± 10 parts per million may be used.
  • Techniques have come into common use since the above frequencies were defined which use very high sampling frequencies, typically more than 100 times higher than 48 kHz. Where such sampling frequencies appear at an interface, only multiples representing some factor of two of the sampling frequencies identified in AES5 should be used.

Format specifications Explanation of format description terms

Useful references


Books, articles, etc.

Last Updated: 08/10/2021