Powered by
JSPWiki v2.8.2
g2gp 17-01-2009
View PDF
This is version . It is not the current version, and thus it cannot be edited.
[Back to current version]   [Restore this version]

Digital Audio#

This Guide is still in draft format.

Section 1. Introduction to Digital Audio#

Section 2. Creating Digital Audio#

Section 3. Preserving Digital Audio#

Bibliography and Further Reading#

Knight, G. & McHugh, J. (2005) Preservation Handbook: Digital Audio. AHDS. http://www.ahds.ac.uk/preservation/audio-preservation-handbook.pdf

Section 1. Introduction to Digital Audio#

1.1 What is Digital Audio?#

As with digital video, digital audio files have become far easier to create over the last ten years. Although digital video has perhaps found more applications in archaeology, digital audio files are often created as a component of projects looking to record oral histories or to recreate 'archaeological sounds', either through the modern reconstruction of archaeological musical instruments or through the recording of sounds within archaeological contexts such as reconstructed - physically or virtually - churches, henges or theatres. A number of examples of the latter exist in the online Internet Archaeology journal (e.g. Thomas 2011).

As with the Digital Video guide, this guide aims to address the issues involved in the creation and preservation of 'born digital' audio files and will not aim to cover the creation of files from analogue originals (although many of the issues discussed here equally apply). The digitisation of analogue audio files is covered in detail in the JISC Digital Media guide 'Audio: Digitising analogue media' [1]. In addition to the extensive material available on the JISC Digital Media site, this guide also draws heavily from a number of other key guides on preserving digital audio files, namely the AHDS 'Preservation Handbook: Digital Audio' (Knight & McHugh 2005), JISC's 'Significant Properties Testing Report: Audio Recordings' (Knight 2010) and IASA Technical Committee's 'Guidelines on the Production and Preservation of Digital Audio Objects' (2nd ed.) (Bradley 2009). A detailed introduction to the technical aspects of digital audio files can also be found in the JISC Digital Media guide 'An Introduction to Digital Audio'[2].

1.2 Current Issues and Concerns#

Again, as with digital video data, digital audio files may be large when created/stored in uncompressed formats and informed decisions need to be made when deciding when and how lower quality files are created. Another issue that is again similar to digital video is that the range of digital audio files consist of a mix of container fomats and codecs which again emphasise the importance of detailed technical metadata in successfully identifying and working with audio files. Metadata also plays a key role with audio files in documenting the file's creation process and contents (e.g. names and dates of interviews, locations, etc.) as these elements may not be as apparent as in similar video files.


[1] http://www.jiscdigitalmedia.ac.uk/audio/docs/category/digitising-analogue-media
[2] http://www.jiscdigitalmedia.ac.uk/audio/advice/an-introduction-to-digital-audio

Section 2. Creating Digital Audio#

2.1 General Considerations#

The process and considerations involved in creating digital audio files are fully discussed in a number of JISC Digital Media documents, specifically:

  • 'Choosing a Digital Audio File Format'[1]
  • 'Basic Audio Editing'[2]
  • 'Choosing a Digital Audio File Format'[3]

As discussed in Section 1, the quality of the file is the key consideration when creating digital audio files. In order to ensure that the file being created is of a suitable quality and fit for purpose, data creators should be aware of :

  • the relationship between audio container or 'wrapper' formats and the codecs that they contain as well as their respective capabilities.
  • the fact that, as a component of the codec, audio formats can feature both lossless and lossy compression or come in uncompressed formats. Data creators should be aware of how these affect both data quality and file size at at what stages in the work flow these options will (if at all) be used.
  • that many files allow metadata to be embedded within the container format and that data creators may want to utilise this functionality to record important aspects of the data creation process.
  • which rights exist (e.g. copyright or moral right). This is particularly relevant to interview situations and data creators should be aware of any ethical implications of their work. In all cases clearance/permission should be sought from those involved in the recording. These issues are discussed in the JISC Digital Media document 'Copyright and Other Rights for Creating Time-based Media Resources'[4].

As a general guide - and one that is repeated throughout these Guides - it is advised that data creators create (and archive) audio data at as high a quality level as is available and in an uncompressed format. From this master dataset, lower filesize/quality datasets can be derived via format migratation or downsampling.

2.2 File Formats#

FormatProperties/TechnologiesDescriptionRecommendations
Waveform Audio (.wav)A widely used and openly documented (though proprietary) container format developed by Microsoft and IBM.The .wav format is commonly used to store uncompressed audio (as PCM) but can also contain audio in a variety of lossy codecs such as MP3. The format can also be tagged with metadata and can embed metadata in the XMP format.Suitable for preservation depending on the codec used.
Audio Interchange File Format (.aif, .aiff)Proprietary format developed by Apple and similar to the .wav format.As with .wav, the .aiff is predominantly used for storing uncompressed PCM audio files. The format is also capable of storing metadata along with audio in other compressed codec formats.Suitable for preservation in uncompressed format.
Sun AU (.au)A format developed by Sun for Unix systems.As with .wav and .aif files, Sun AU files tend to be large and of high quality but are not widely supported outside the UNIX community. Although predominantly using PCM encoding, .au files can support a number of of other codecs.Not widely supported outside the UNIX community so therefore not suitable for preservation.
Advanced Audio Coding (.aac)An ISO standard format based on MPEG-2 and MPEG-4 formatsThe format is widely supported by a number of common devices (Wii, Playstation, iPod, iPhone amongst others) and is designed to be the successor of the MP3 format. AAC format data may be commonly packaged in MP4, 3GP, ADIF and ADTS container formats.Suiatble for dissemination but not for preservation.
Broadcast Wave Format (BWF) (.bwf .wav)A specification of the European Broadcasting Union and an extension of the WAV format[5].Broadcast Wave Format consists of uncompressed audio that extends the WAV format with an additional 'chunk' for metadata. Although based on the same format, compatibility between WAV and BWF files can be problematic, particularly when migrating WAV to BWF.Suitable for preservation.
Ogg Vorbis (.ogg)A non-proprietary open format developed by Xiph.org[6].The Vorbis codec has been developed as a completely open alternative to the MP3 format and primarily as a dissemination format (the Ogg wrapper format can also contain codecs other than Vorbis e.g. FLAC).Suitable as a dissemination format.

MP3 / MPEG-1 Audio Layer -III (.mp3)A widely accepted format that can be played on most platforms. Several patents cover the format, although a free licence is granted to non-profit organizations. Suitable for distribution, but ill equipped for preservation.
MP3: MPEG-1 Audio Layer 3, commonly shortened to “MP3” is an audio encoding format that uses a lossy compression algorithm to reduce the amount of data required to reproduce an audio recording. The compression is based upon the principle of 'perceptual coding' which discards or reduces accuracy of sections of sound that are considered beyond the hearing of most people. An MP3 audio file frequently contains an additional ID3 metadata component.

m4p – A proprietary version of AAC in MP4 with Digital Rights Management developed by Apple for use in music downloaded from their iTunes Music Store.

Quicktime (.qt, .mov)Apple proprietary streaming Codec. Not suitable for preservation

Real Audio (.ra, .ram)Real Networks proprietary streaming Codec. Not suitable for preservation.

Windows Media Audio (.wma)Microsoft proprietary streaming Codec. Not suitable for preservation.

FLAC (.flac)File format for the Free Lossless Audio Codec, a lossless compression codec
FLAC: FLAC15 (Free Lossless Audio Codec) is a non-proprietary file format for the storage of audio as lossless, compressed data. It was developed and maintained by Xiph.Org Foundation, which promote it as a lossless replacement to the popular MP3 format. FLAC refers to an encoding format for the storage of audio streams and a ‘Native Flac’ container format for the storage of disparate information. A FLAC encoded audio stream may be embedded in several container formats, including Native Flac and Ogg.

.

vox – the vox format most commonly uses the Dialogic ADPCM (Adaptive Differential Pulse Code Modulation) codec. Similar to other ADPCM formats, it compresses to 4-bits. Vox format files are similar to wave files except that the vox files contain no information about the file itself so the codec sample rate and number of channels must first be specified in order to play a vox file.

(.wma) – the popular Windows Media Audio format owned by Microsoft. Designed with Digital Rights Management (DRM) abilities for copy protection.

http://en.wikipedia.org/wiki/Audio_file_format

Notes on formats / Future Directions#

.wav and .aif are uncompressed audio files and are therefore the only suitable formats for delivery and preservation. Whereas .wav is the default format for MS Windows, .aif is the equivalent on a Mac. As we work on Windows PCs, here at the ADS, .wav is the preferred delivery format.

As well as a raw audio file (.wav, .aif, .au), depositors should also supply a compressed version of their audio files that we can use for web delivery. There are a wide number of different formats that we can accept for delivery. We can create these ourselves if need be but this will obviously require time and money!

In situations where the depositor supplies only the compressed version of their audio files (for example .rm, .wma, .mp3, .ogg), they should be strongly encouraged to send us uncompressed originals as well. It is possible that the device they recorded with saved directly to a compressed format. If for whatever reason the depositor can not give us uncompressed files, we can use a free conversion utility to convert the files back to an uncompressed .wav format. It should of course be noted that by converting back to .wav we will not be regaining any of the information and quality from the uncompressed original, we will just be creating a file that we can refresh alongside other preservation files in the future.

A number of open source audio formats are being developed by Xiph.org (e.g. FLAC, Speex, Ogg Vorbis, etc.) and it may be that these become more widespread and therefore more suitable for us to accept as deposit and / or archival formats. FLAC (Free Lossless Audio Codec), as a file format that uses lossless compression (c. 40-50% reduction of file size) is the most likely of the Xiph.org formats to be of interest to us as a possible archival format for large audio files/collections. In terms of disseminations formats it is likely that MP3 will be superceded by the AAC/M4A developed by the MPEG group and currently used by Apple's iPod (see AHDS Moving Pic and Sound, p51 and Wikipedia: Advanced Audio coding).

Other formats#

When compressing audio for web delivery, we have to make decision about which codec (compressor/decompressor) to use. Compressing audio files will produce some loss of quality. Lossless codecs can be found (for example SHN (Shorten) and Apple Lossless) but they seem to only half the file size whereas MP3 can reduce a file to about a tenth of the original size. This inevitable loss of quality can be minimised by use of the correct codec. Different compressors can give very different results and choice of compressor may depend on the type of audio that has been recorded. MP3 may be a popular option for compressed digital audio but is designed primarily for music and is not designed to be streamed. Audio files deposited with the ADS are likely to be recordings of the human voice for oral history projects. As the human voice has a relatively small range, it is a good idea to compress it with a codec designed specifically for this purpose (for example the Quicktime codec Qualcomm Pure Voice), or the open source and free Xiph.org speex.

  • Both compressed and uncompressed audio files
  • All copyright issues clarified and documented
  • Full data documentation and transcriptions if appropriate
  • If depositor has used a file format that contains metadata, they should inform us of this so we can extract it and preserve, or ensure it's preserved in any future versions
  • Extract metadata from files if appropriate

[1] http://www.jiscdigitalmedia.ac.uk/audio/advice/choosing-a-digital-audio-file-format
[2] http://www.jiscdigitalmedia.ac.uk/audio/advice/basic-audio-editing
[3] http://www.jiscdigitalmedia.ac.uk/audio/advice/choosing-a-digital-audio-file-format
[4] http://www.jiscdigitalmedia.ac.uk/crossmedia/advice/copyright-and-other-rights-for-creating-time-based-media-resources
[5] http://www.ebu.ch/fr/technical/publications/userguides/bwf_user_guide.php
[6] http://www.xiph.org

Section 3. Preserving Digital Audio#

  • Archiving # data
    • deciding what to archive
      • Selection and retention
      • preservation intervention points / file and data lifecycles (specific to guide, will also be covered generically)
    • deciding how to archive
      • archiving strategies (migration (to new format, to 'basic' format), emulation, refreshment)?
      • significant properties
      • file types
    • Metadata and Documentation
      • project level
      • file level
      • standards specific to #
    • Structuring your archive

File Formats for archiving

JISC DM uncompressed formats http://www.jiscdigitalmedia.ac.uk/audio/advice/uncompressed-audio-file-formats

WAV or BWF Some projects may require an open standard format. In these cases the suggested format is AIFF. (from JISC creating... document)

3.3 Metadata and Documentation#

ElementDescription
Software used to create/encode (if applicable)
Bit rate (kbps)
sampling frequency range (KHz) if applicable
Codec used (where appropriate)
Length of recording (mins and secs).
Copyright clearances (very important for audio files, especially oral history).
Transcriptions of interviews etc (where appropriate).

Embedded metadata "indicating its purpose, who created it and when, the actions that have been performed on it, associated rights and other information." (Knight 2010)

http://www.iasa-web.org/tc04/basic-metadata

  • Copyright
    • specific copyright considerations for each guide.

  • Case study/studies


Digital Audio: Bibliography#

Bradley, K. (ed.) (2009) Guidelines on the Production and Preservation of Digital Audio Objects. Second edition. IASA Technical Committee www.iasa-web.org/tc04/audio-preservation(info)

Thomas, D. (2011) 'An Investigation of Aural Space inside Mousa Broch by Observation and Analysis of Sound and Light'. Internet Archaeology 30. http://intarch.ac.uk/journal/issue30/thomas_index.html

Knight, G. & McHugh, J. (2005) Preservation Handbook: Digital Audio. AHDS. http://www.ahds.ac.uk/preservation/audio-preservation-handbook.pdf

Knight, G. (2010) Significant Properties Testing Report: Audio Recordings. JISC. http://www.significantproperties.org.uk/testingreports.html