Media Servers

How Audio Files Work: Codecs and Containers

This article explains common audio file extensions, codecs, and containers pertinent to consumer home theater and music projects.

Looking for an article on video codecs and containers? Checkout Video Codecs, Containers, and File Extensions

Table of Contents

Audio Codecs and Containers are frequently ignored and easily mixed up when it comes to the digital recording of audio. The important questions around "which file type to use" should not revolve around file extensions. It is the underlying audio codec and the codec container type that are the critical factors. Focusing solely on file extensions (as most people do) without understanding what they represent leads many end users down a path of frustration with their digital collections of audio content (particularly music and movie soundtracks).

Consumers have been trained to think of file extensions in absolute terms and indicators of a file's content. Yet, the reality is file extensions are arbitrary, and in some cases the same file extension may represent more than one type of data or associated application. Multimedia files can be particularly puzzling.

This article is part of a series that seeks to demystify the most important aspects of digital multimedia files. It delves into the most common audio codecs and containers, file extensions, lossy and lossless audio compression, and operating system compatibility. A cursory discussion of metadata is also included. Several other articles complement this one.

Multimedia Codecs, Containers, Filenames, and Metadata provides a higher level, broader overview compared to this current article.

Video Codecs, Containers, and File Extensions mirrors this article's discussion, with a focus on video containers, codecs, and common video, audio/visual file extensions.

Multimedia Metadata Deep Dive takes a deeper dive into the important, anecdotal subject of metadata.

How to Use This Article

Frequently asked questions and how to use this article to find the answers.

Which Audio Codec is this File Extension?

File extensions indicate a file's LIKELY content type. You may want to match an audio file extension to a content type, or correlate with Audio Codecs and/or Containers.

Which File Extension should I save my files with?

When someone asks a question such as, "Should I save my music files as MP3's?" What they are really asking is, "Should I record digital copies of my files using an MP3 codec?"

The answer depends on your needs and goals. Determining which digital audio codec to record source content onto depends on factors such as how many channels need to be recorded to the file (e.g. stereo or 5.1 surround sound). Some codecs support multi-channel surround and some don't. When examining more advanced codecs, a few will retain audio objects; most will not. A few codecs can record more than eight (8) simultaneous channels; most cannot. If you have specific or complex needs, it pays to be diligent in assessing how to go about recording music and other audio sources prior to adopting a particular approach.

Which Audio Codec do I need?

This depends on your goals. The single most important factor is the required number of audio channels. Check out the table correlating audio codecs with their maximum number of channels under Audio Codecs.

Topics Not Covered

Certain codecs and/or containers are not discussed in this article, including:

  • Compact-Disc Audio (.cda)
  • Super Audio CD (SACD)
  • DVD-Audio
  • Direct Stream Digital (DSD)
  • Real Audio (.ra)
  • MPEG-D (see MPEG Audio Standards)

Audio File Extensions

Audio file extensions should indicate the audio codec and/or container format of a given file.

A file extension symbolizes the type of data a file contains. It's certainly possible for the data type to be something other than what is expected, but standardized file extensions are incredibly useful. They allow both humans and computers to rapidly identify whether or not a particular file is a type they're familiar with.

Audio file extensions are typically associated with containers. Under some circumstances, a file is a codec-encoded audio file not encased inside a container.

Why Use Containers?

If it's possible for an audio codec to stand-alone as a file, then why bother placing one inside a container to begin with?

Constainers serve several useful purposes.

  • Most containers allow the incorporation of metadata. Since most codecs to not use metadata, placing the codec inside a container improves flexibility in managing the audio content.
  • Some containers allow more than one audio file inside of them. For example, a combination of a mono-channel track, stereo-channel track, and multi-channel track can all be stored inside a single container type that supports multi-channel segmentation of audio.
  • Many containers allow combining video and audio content in the same file.

Which digital file format should I record or save my music to? This is a common question from consumers. Unfortunately, there is no simple answer. Certain file types are the most frequently encountered, but just because other people use them doesn't mean they are the ideal solution for you. It is worthwhile to examine your particular situation and make an informed choice.

MP3 is by far the most ubiquitious digital music file format. However, that's the case not because it is the best, but due to how the use of digital music has evolved over time. What makes MP3 so popular?

  • Nearly universal decoder support
  • Small file sizes
  • Exceptionally common file format for downloads
  • Good quality audio on most devices
  • Average person does not perceive defects in audio quality from compression
  • Longevity; MP3 has been around for over 20 years

A primary historical reason for MP3's popularity was its small file size, which facilitated storing digital music libraries on a single storage volume. When MP3 was released, its only viable competitor at the time was WAVe. While Wave files had (and still have) markedly superior audio quality, this comes at the cost of huge file sizes. In contrast, MP3 offers a very good audio copy at a fraction of the file size. This is due to MP3's lossy file compression, versus uncompressed Wave files.

FLAC is modern, popular lossless digital file format for music. Its popularity stems largely from the fact it is lossless (MP3 is a lossy audio codec/format), and like MP3, FLAC is open source, meaning you may use its codec for free. FLAC is widely supported, but is not universal like MP3. If you use Apple products exclusively for recording and playing back music, you may find Apple's lossless audio format to be a better fit.

The table below is a comprehensive overview of the most common audio codecs and containers, sorted by year of introduction.

Common Audio File Extensions, Codecs, and Containers (by Year)
File Extension Full Name Type Lossy/
Lossless
Meta
data
Year Operating System Support
Win Lin Mac iOS And
.aif Audio Interchange File Format (AIFF) Container Yes 1988 Yes No Yes No No
.aifc AIFF - Compressed (AIFF-C) Container Yes 1991 Yes No Yes No No
.riff Resource Interchange File Format (RIFF) Container Yes 1991 Yes Yes Yes No No
.snd | .au Sun Microsystems Audio File (Au) Container Lossless No 1991 No Yes No No No
.wav Wave Container Yes 1991 Yes Yes No No Yes
.ac3 Dolby Digital (AC-3) Codec Lossy No 1991 Yes Yes1 Yes Yes Yes
.dts Digital Theater Systems (DTS) Codec Lossy No 1993 No Yes1 No No No
.mp1 MPEG-1 Audio Layer I (MP1) Codec Lossy No 1993 Yes Yes No No No
.mp2 | .mpa MPEG-1 Audio Layer II (MP2) Codec Lossy No 1993 Yes Yes No No No
.mp3 MPEG-1 Audio Layer III (MP3 gen1) Codec Lossy Yes 1993 Yes Yes Yes Yes Yes2
.ogg Ogg Container Yes 1993 Yes Yes No No Yes
.shn Shorten Codec Lossless No 1993 No No No No No
.wav Exchangeable Image File Format (EXIF) Container Yes 1995 Yes Yes Yes No No
.mp3 MPEG-2 Audio Layer III (MP3 gen2) Codec Lossy Yes 1995 Yes Yes Yes Yes Yes2
.ra Real Networks RealAudio Codec Lossy No3 1995 No No No No No
.aac Advanced Audio Coding (AAC) Codec Lossy No 1997 Yes Yes Yes Yes Yes
Advanced Audio Coding Plus (AAC+) Codec Lossy No 1997 Yes Yes Yes Yes Yes
.adts Audio Data Transport Stream Container No 1997 Yes No Yes No Yes
.mpc Musepack Codec Lossy Yes 1997 Yes Yes Yes No No
.wv WAVPack Codec Either4 Yes 1998 No Yes No No No
.amr Adaptive Multi-Rate (AMR) Codec Lossy No 1999 No No Yes No Yes5
.tta True Audio (TTA) Codec Lossless Yes 1999 Yes No No No No
.wma Windows Media Audio (Lossy) Codec Lossy Yes5 1999 Yes Yes No No No
.asf Advanced Systems Format (ASF) Container Yes 2000 Yes No No No No
.ape Monkey's Audio (APE) Codec Lossless Yes 2000 Yes No No No No
Vorbis Codec Lossy Yes 2000 Yes Yes No No No
.flac Free Lossless Audio Codec (FLAC) Codec Lossless Yes 2001 Yes Yes Yes Yes Yes
.mka Matroska Container Yes 2002 Yes Yes No No Yes
.m4a MPEG-4 Part 14 (MP4) Container Yes 2003 Yes Yes Yes Yes No
.wma Windows Media Audio (Lossless) Codec Lossless Yes 2003 Yes No No No No
.wma Windows Media Audio Pro Codec Lossy Yes 2003 Yes No No No No
.3gp 3GPP Container No 2004 Yes No No Yes Yes
.alac Apple Lossless Audio Codec (ALAC) Codec Lossless Yes 2004 Yes No Yes Yes No
.mxf Material eXchange Format (MXF) Container Yes 2004 Yes Yes Yes Yes Yes2
.caf Core Audio Format (CAF) Container Yes 2005 No No Yes Yes No
.dts DTS-HD Master Audio (DTS-HD MA) Codec Either No 2005 No Yes1 No No No
.als Audio Lossless Coding (ALS) Codec Lossless No 2006 No No No No No
.ac35 Dolby Digital Plus (DD+) Codec Lossy No 2006 Yes Yes1 Yes Yes Yes
.sls Scalable Lossless Coding (SLS) Codec Lossless No 2006 No No No No No
.ac3 | .thd Dolby Digital TrueHD Codec Lossless No 2008 No No No No No
.webm WebM Container Lossy Yes 2010 Yes8 No No No Yes
.opus Opus (Ogg Opus) Codec Lossy Yes 2012 No Yes Yes9 Yes9 Yes2,10
.ac3 | .thd Dolby Digital TrueHD + Atmos Codec Lossless No 2014 No No No No No
.dts DTS:X Codec Either No 2015 No No No No No
.dts DTS:X Pro Codec Either No 2019 No No No No No

Codecs and Containers

Codecs and containers play different, but equally important roles in sharing multimedia. The most important factor is whatever device you wish to read/play multimedia content on must use the same codec and container as the device that created the content in the first place.

A codec is a device or computer program that encodes and/or decodes digital data using mathematical algorithms. Codecs facilitate the storage, retrieval, and sharing of information. Most codecs provide compression and decompression options to reduce the original size of a file, reducing the amount of information that must be stored (encoded) and retrieved (decoded).11 Compression is a separate concept from the lossiness of a codec.

Lossy codecs use data compression techniques to reduce the storage footprint of data even further. They are called "lossy" techniques because some audio data is inherently lossed in the process. Most of these algorithms use a form of fuzzy logic, though more recent and modern codecs are increasingly likely to include algorithms based on machine learning or artificial intelligence. The Opus audio codec is a good example of this.12

An audio codec is a format or structure that provides storage and transmission capabilities for audio content.13 Sometimes described as an audio coding format or audio compression format, an audio codec may be described as an algorithm, but in reality it more closely resembles a schema: a system or plan that describes how information shall be represented and recorded within a file.

Approximate Logic is an emerging data modeling technique with the potential to bestow improvements to audio codecs, particuarly with regards to the ratio of audio quality to file size. You may think of approximate logic methods as a sort of advanced fuzzy logic.

Audio codecs make the storage, sharing, and streaming (transmission) of audio data more efficient and sometimes faster compared with sharing raw audio. They accomplish this feat through various techniques including data compression and lossy algorithms. Some codecs allow bundling multiple streams of data together into a single collection of digital data; for example, encoding multiple audio channels into a single data stream that may then be decoded by another device or process.

Lossy storage formats always utilize data compression, but compressed data is not necessarily lossy.

A container is a digital storage package. Containers hold one or more distinct multimedia streams. Each stream is normally encoded (via a codec). The container keeps track of information about each stream it is holding by following a schema, much like a database. Just like codecs, a device that intends to restore the data inside a container must understand how the particular container is organized; it must be capable of reversing the containment process. Likewise, a receiving device will device will also need to be able to be capable of using the same codec(s) that any multimedia data inside the container was encoded with, in order to decode the actual data streams and restore them to pure audio and/or video.

Most containers support contextual metadata. Most codecs do not support contextual metadata natively.

Ultimately, stored or transferred audio data must be recreated in its true (digital) form: PCM.

All Codecs Lead to PCM

When it comes to digital audio on virtually any device, all these codecs and containers boil down to accomplishing one thing: converting sound to or from Pulse Code Modulation (PCM).

PCM is the equivalent of "raw audio" to a computer. When you record sound on a digital device, its native format will be PCM (or LPCM). If you want to play it back, it must be converted to PCM first or the device will not be able to reproduce the analog signals us humans consider to be "sound." The key is not how you compress, finesse, store, and retrieve the sound data, it's when and how these conversions to and from PCM occur.

Audio recorded to a DVD or Compact Disc, and that used in digital telephony are all PCM streams.

Raw Audio Codecs: PCM, LPCM, and ADPCM

As mentioned above, Pulse Code Modulation or PCM based codecs digitally represent analog signals. You may think of PCM as the most pure form of audio signals to a computer.

Most of the time, the differences between variations in PCM versions (e.g. PCM, LPCM, DPCM, and ADPCM) amount to "splitting hairs," per se. What is important to understand is that an identical decoding standard must be employed to match the particular codec used to encode audio content to begin with.

PCM

PCM refers to a process of transforming sound waves between an analog waveform human ears interpret as audio and a digital representation computing devices are capable of storing and manipulating, called a pulse code modulated waveform.

PCM is the "gold standard" of digital audio, because virtually all audio streams are ultimately converted into PCM. You may think of PCM as the most pure form of audio possible on a digital device. It is lossless, uncompressed, and requires substantial bandwidth. In fact, the primary reason for the existence of audio compression and lossy codecs are due to the transmission bandwidth required to stream pure PCM. Pure PCM files require considerable storage capacity when the data is in a file; particularly if an audio stream contains multiple independent channels in the same recording.

LPCM

What are Pulse Code Modulation (PCM) and Linear Pulse Code Modulation (LPCM)?

LPCM stands for Linear Pulse Code Modulation. Most PCM data is actually encoded as LPCM, and with regards to computers the terms are often used interchangeably. For the purpose of this discussion, their differences are not important. You may think of LPCM as a sort of slightly more refined version of PCM.

Also known as PCM Plus or PCM+, Linear Pulse Code Modulation (LPCM) is simply a specific (and common) type of PCM. Most devices supporting PCM also support LPCM. Audio sent through HDMI cables and written onto Blu-Ray DVDs is normally LPCM.

PCM and LPCM are the most common - and universal - big dogs on this camp. ADPCM is mentioned primarily due to Microsoft's and Apple's support of IAM (ADPCM) codecs; though they are not widely utilized or known, and ultimately the data is distilled down to - you guessed it - PCM.

AU Audio, WAV, AC3, AIFF, AES3, RF64, MP3, and M4A codecs all decode to LPCM. DVD and Blu-ray audio is recorded as LPCM.

ADPCM

Adaptive Differential Pulse Code Modulation or ADPCM is an oddball. A variation on the PCM concept, ADPCM refers to a group of codecs that act like containers. There are different codecs for ADPCM implementations on different operating systems. These codecs are proprietary implementations of PCM. One of the most common is IMA-ADPCM.14 Another is WAVEFORMATEX; created by Microsoft.

Conceptually, ADPCM itself is simply raw PCM audio data with a structural wrapper around it which defines blocks of PCM data. ADPCM codecs are lossy and compressed, and follow a pre-determined structure that defines the PCM content inside the file. ADPCM codecs are typically only found on standard computer operating systems (e.g. Linux, Windows, MacOS), while mobile operating systems such as iOS and Android usually don't support them. Mobile devices rarely if ever use it, finding other formats better suited to their needs, such as AMR. The latter incorporates not only audio compression, but is also designged to be robust in dealing with flaky mobile connectivity issues. Either way, ADPCM is a lossy codec typically limited to use in short bursts of PCM audio content, such as speech-oriented applications (e.g. VoIP transmissions). Putting all these facts together, you'll find ADPCM is not used very often.

Multi-Channel: Codec or [Codec] Extension?

It's easy for the average person to feel overwhelmed trying to figure out multi-channel audio terminology as it pertains to audio files, file extensions, and how these correlate to multi-channel sound. Which technologies are audio codecs? Which are features of a codec? Which are features of a device? This confusion is frequently pertinent to complex surround sound/multi-channel codecs. Their nomenclature often doesn't help. For example, while several Dolby Digital and DTS formats are codecs, most anecdotal technologies associated with them are not (e.g. Dolby Atmos).

Multi-channel format support is device dependent, not just operating system dependent.

Why is this important relative to audio file extensions? If you're aspiring toward a particular type of surround sound recording in an audio file, it's important to understand which file types may or may not be capable of supporting the type of audio you're interested in. While that subject gets complicated pretty easily. For example, true 6.1 discrete channel output vs. matrixed (pseudo) 6.1 output are both "6.1" but there is a difference between the actual sound recordings, file size, and the PCM audio sent to your audio output device, such as a home theater A/V receiver.

Dolby Laboratories and DTS Codecs

Dolby and DTS are the most frequently encountered multi-channel codecs. Their audio codecs support up to 32 channels of sound (though most devices limited to 14 channels at the upper end). Only six (6) audio file codecs support more than eight (8) discrete channels, and of those only two (2) enjoy native operating system decoding support: AAC (MPEG-4) and EAC3 (Enhanced AC-3).

Dolby Atmos and DTS:X

Dolby Atmos and DTS:X are competing so-called "immersive" surround sound techniques that eschew traditional directed audio channels in favor of audio "objects" which are basically

Dolby Atmos tends to be a confusing subject when it comes to understanding how it relates to audio codecs. The fact is, there is no codec for Atmos specifically. Likewise, the same is also true of DTS:X, DTS:X Neural, and Auro-3D. Yet, to make matters more confusing, those all require codecs to function properly. So then, how are they codecs but not codecs? The answer is they require what Dolby and DTS refer to as extensions; which are functions added on to existing codecs.

A simplified way to look at this is to think of codecs having versions. You may have let's say a Dolby TrueHD incorporated into the decoder inside your A/V receiver, however if the receiver's codec does not also contain the Dolby Atmos extension - which sits on top of the TrueHD codec - then you won't be able to play native Atmos audio content. In that case, your receiver would simply ignore the Atmos content it doesn't understand.

Older codecs will ignore newer codec formats they don't understand. Under some circumstances, the old codecs will still be capable of playing some sound. Whether or not that is the case depends on how the content was encoded rather than which decoder you're using to attempt to listen to the content.

Dolby AC-4

Dolby AC-4 is the most recent lossy audio codec from Dolby labs. Released in 2018, it is likely to begin trickling into the audio/video world in earnest in 2020. From a structural viewpoint, AC-4 in some ways resembles a container format because it is organized around the concept of audio objects (represented as frames), though streams consisting of channelized audio content are embedded inside the frames.

It seems very likely Dolby will utilize AC-4 as a stepping stone to move toward codecs that package audio as audio objects only, a-la Auro's Auro-3D codec (2011), which proved the concept is favorable. However, content authors and streamers should expect substreams using traditional channelized audio will continue to be in high demand for quite some time. Less capable devices with limited audio channel outputs or devices with limited bandwidth may prefer to receive audio data encompassing minimal channel encoding due to their hardware and presentation limits, potentially discarding object oriented and surround sound channelized bitstreams to conserve bandwidth or because the form factor means it's pointless to transfer such data types.

A comprehensive discussion of multi-channel audio formats is beyond the scope of this article.

For a deeper discussion of multi-channel audio codecs, see Epic Battles of Home Theater: Dolby, DTS, and Auro.

Auro: Auro-3D

Auro-3D or simply Auro is the latest entrant into the surround sound codec rodeo. Sporting a maximum of 32-channels, Auro is the first audio codec that relies exclusively on the latest concept in multi-channel audio content delivery: audio objects.15 To date, there is no specific file extension associated with Auro-3D.

Audio Codecs

An audio codec is a format or structure that provides storage and transmission capabilities of audio content.13 Sometimes described as an audio coding format or audio compression format. Audio codecs are not just algorithms. They are also schemas: systems or plans describing how information shall be recorded (encoded) and presented (decoded). Modern encoder and decoder codecs are almost always completely different algorithms. It is a strategy very similar to cybersecurity encryption key handling.

Looking for details on Audio Containers?

Audio codecs process, store, retrieve, and (in some cases) transmit audio based content including music, speech, and other sounds within the range of human hearing (~20hz - 20Khz). The chart below describes the most frequently encountered audio codecs. While not exhaustive, it is comprehensive.

Audio Codecs
Codec Notation Channels Decoder Native Operating System Support
Win-10 Linux macOS iOS Android
Advanced Audio Coding (AAC)16 AAC 4817 Yes Yes Yes Yes Yes
Advanced Audio Coding Plus (HE-AAC)18 AAC+ 4817 No No No Yes Yes
Dolby Digital AC-3 (Audio Coding 3) AC-3 5.1 Yes Yes1 Yes Yes Yes
Adaptive Differential PCM (ADPCM) ADPCM 2+19 Yes Yes Yes20 Yes No
Apple Lossless Audio Codec (ALAC) ALAC 8 Yes No Yes Yes No
Audio Lossless Standard (ALS) ALS 65,536 No No No No No
Adaptive Multi-Rate Audio Codec (AMR)5 AMR 1 No No Yes No Yes21
Monkey's Audio (APE) APE 8 Yes No Yes22 Yes22 No
Digital Theater Systems (DTS) DTS 5.1 No Yes1 No No No
DTS-HD Master Audio (DTS-HD MA) DTS 11.1 No No No No No
DTS:X23 DTS 11.1 No No No No No
DTS:X Pro24 DTS 32 No No No No No
Dolby Digital Plus (DD+/Enhanced AC-3) E-AC-3 15.1 Yes Yes1 Yes Yes Yes
Free Lossless Audio Codec (FLAC) FLAC 8 Yes Yes Yes25 Yes26 Yes
MPEG-1 Audio Layer I (MP1) MP1 2 Yes Yes No No No
MPEG-1 Audio Layer II (MP2) MP2 2 Yes Yes No No No
MPEG-1 Audio Layer III (MP3 gen1) MP3 2 Yes Yes Yes20 Yes Yes27
MPEG-2 Audio Layer III (MP3 gen2) MP3 5.1 Yes Yes Yes20 Yes Yes27
Musepack MPC 8 Yes Yes Yes No No
Opus Interactive Audio (Opus) OPUS 255 No Yes Yes9 Yes9 Yes27,28
Pulse Code Modulation (PCM)29 PCM 6553530 Yes Yes Yes Yes Yes
Scablable Lossless Standard (SLS) SLS 65,536 No No No No No
Dolby Digital TrueHD31 THD 832 Yes No Yes Yes No
True Audio TTA n/a Yes33 No No No No
Vorbis Vorbis 255 Yes Yes No No No
WavPack WavPack 16 No Yes No No No
Windows Media Audio (WMA) [Lossy] WMA 6 Yes Yes No No No
Windows Media Audio (WMA) [Lossless] WMA 6 Yes No No No No
Windows Media Audio Pro (WMA Pro) WMA Pro 7.1 Yes No No No No

Multi-Channel Codecs

These codecs are capable of encoding and decoding multiple audio channels in a single stream. The encoding process of these codecs can be viewed as a form of multiplexing. For the purpose of this discussion, "multiple" means more than two (2) audio channels (stereo).

Multiplexing is the process of merging multiple sources of audio, video, and/or closed captions into a single data stream. Sometimes referred to as muxing or muxing streams, these terms all mean the same thing. Muxing is simply a slang term for multiplexing.

Do I Need More Than 8 Audio Channels?

Astute readers may have noted PCM only supports a maximum of eight (8) distinct audio channels. Therefore, is there any point to using codecs that support more than eight (8) channels?

At the moment, this is a limitation of the PCM standard. However, there is no reason to be alarmed; at least not yet. The most common conduit for audio information to Hi-Fi equipment is still HDMI, which - because it uses LPCM - is limited to 8 simultaneous audio channels.

That doesn't mean you shouldn't consider using audio codecs that support more than eight (8) channels. Audio equipment gets around the current 8-channel bottleneck by delivering audio data to a device faster than it can be played or through the use of non-channel based content, such as what are called sound "objects." When audio data is streamed rapidly to a device, the excess data is buffered (stored in memory) by the device until needed. This is one reason why PCM has remained mired in its 8-channel limit while other portions of the audio industry have evolved (e.g. Dolby TrueHD + Dolby Atmos). There isn't yet a problem from a user experience perspective.

Subjectively, I don't see any reason for any of this to change anytime soon. HDMI for example continues to gain more throughput speed with each new standard. No matter how many audio channels may be supported by equipment in the future, it seems exceptionally unlikely that it will ever become a problem for the equipment to pre-load or buffer content much faster than it is required to play that content via analog speakers. Therefore, I see this limitation of PCM as a non-issue.

Stand-alone audio files can have their own file extensions don't have to be inside a container.

Popularity: Popular
Support: Good
Type: Lossy
Stand-alone: .aac
Genre: Music

AAC (Advanced Audio Coding)

Launched in 1997 as the successor to MP3, Advanced Audio Coding (AAC) offers smaller files and faster encoding while maintaining as-good-as or better audio fidelity. Although widely supported, AAC remains relatively unknown to most consumers.

Over time, the original AAC standard has evolved into a family of proprietary MPEG-2 and MPEG-4 lossy audio codecs. The earliest (MPEG-2) iteration can be thought of as AAC Version 1, as it is the original (main) AAC standard.34 The MPEG-4 reference is also commonly also referred to as AAC, but in order to clarify the version it may be referred to it as AAC Version 2. The official names for the MPEG-4 main AAC version are AAC-LC (where "LC" means "Low Complexity") and/or AAC Profile. Other forms of MPEG-4 AAC protocols include ALS (MPEG-4 Audio Lossless Coding), SLS (MPEG-4 Scalable Lossless Coding), and ADTS (Audio Data Transport Stream).

Other sections describe portions of the AAC family in more detail, such as HE-AAC, ALS, SLS, and/or the ADTS file format.

Today, the term AAC has several connotations, and it is often difficult to know exactly which version of AAC someone is talking about unless they clearly distinguish it, such as AACv1, AACv2, etc. For the most part, it doesn't matter as most MPEG-4 based decoders handle MPEG-2 AAC codecs as well.

The AAC-LC term is sometimes used explicitly to clarify discussions between AAC and AAC+ (another name for HE-AAC). Making matters potentially more confusing at times, AAC-LC (also known as AAC version 2) refers to the default codec of the revised AAC standard introduced in MPEG-4, while the original AAC specification may be referred to as AAC version 1 to denote the original AAC specification defined under MPEG-2. When simply the term AAC is used, it could refer to either version 1 or version 2 (AAC-LC). In a similar vein, AAC+ has both a version 1 (MPEG-2) and a version 2 (MPEG-4), although they are both routinely referred to as simply HE-AAC (without a specified version number). Unless someone is specific in a discussion regarding which version of AAC or HE-AAC (AAC+) they are talking about, it could be either the MPEG-2 or MPEG-4 variety.

A Few Words on AAC File Extensions
While AAC may be found as a stand-alone file type (.aac), as previously mentioned it may also be stored inside various container types. As a file format, AAC is a bit more complicated than simply a file extension. The profile type must also be derived, and the underlying format determined before decoding is possible. For most users, these details don't matter. However, if your decoding application is not reading an .aac file properly, the cause could be an unsupported AAC profile. Virtually any MPEG-4 compatible decoder should work.

Though capable of acting as a stand-alone audio stream, AAC streams are often embedded inside audio and multimedia containers. AAC is commonly found inside CAF, Matroska, and MPEG-4 containers such as M4A. AAC audio content bearing the .aac file extension is not necessarily AAC content encased in a (stand-alone) file format. It is possible for an AAC stream to reside inside an MP4, 3GP, ADIF, or ADTS container while bearing the .aac file extension.35 Obviously, this is potentially confusing for playback/decoding applications, which may expect an .aac file to be a raw AAC stream and not wrapped in one of the aforementioned containers. Other container/AAC combinations normally take on the file extension of the corresponding container (e.g. .mpa).

Standards: There are two (2) current root AAC standards. One originally defined under MPEG-2 Audio (Part 7), and the other under MPEG-4 Audio (Part 3, sub-part 4).

For a full cross-compatibility reference of AAC standards, see MPEG Audio Standards.

Popularity: Uncommon
Support: Limited
Type: Lossless
Stand-alone: .alac
Genre: Music

ALAC (Apple Lossless Audio Codec)

Released in 2004, Apple Lossless Audio Codec is a proprietary, compressed 8-channel codec invented by Apple. Arguably one of the best solutions for iOS and macOS platforms, ALAC may be implemented as either a lossy or lossless format. ALAC files can be standalone or stored inside MPEG-4 Audio (M4A) or Apple Core Audio Format (CAF) containers.36

ALAC began rising in popularity within the Apple device community after 2011, when Apple relinquished its licensing requirements and declared ALAC to be an open source platform.37 However, by that time Apple was late to the party and its attempts at making ALAC a global standard have since withered, with FLAC reigning as the most popular lossless digital codec standard due to its ubiquity and nearly universal operating system support.

Standard: Apple Lossless Audio Codec published by Apple, Inc.

Popularity: Uncommon
Support: Poor
Type: Lossless
Stand-alone: .ape
Genre: Music

APE (Monkey's Audio)

Monkey's Audio, often better known by its file extension, APE is a compressed, lossless audio codec. Monkey's Audio is not a very popular audio codec largely due to its very limited operating system support. It is widely supported on Windows, but requires third-party applications on every other operating system. It is virtually unheard of Mac operating system and mobile platforms.

Contributing to Monkey Audio's low adoption rate is the fact it is proprietary. Although it appears to have a liberal licensing policy, though the process is closely held (one must ask the author for prior permission). It's primary functional use-case appears to be compressing raw PCM files (such as lossless WAVe), resulting in lossless but compressed APE files. Monkey's Audio scores higher in data compression tests on the encoding portion and receives relatively negative views with regard to its decompression/decoding aptitude when compared with other lossless audio codecs.38.

Popularity: Common
Support: Good
Type: Lossy
Stand-alone: .ac3
Genre: Home Theater

Dolby Digital (AC-3)

Released in 1991, Dolby Digital AC-3 was the first 5.1 channel "surround sound" codec. Officially known as Audio Coding 3, the AC-3 codec is a proprietary, compressed, lossy audio codec with multi-channel audio coding capabilities (supporting up to 6 channels as 5 full-range channels and one (1) LFE - Low Frequency Emissions - channel). AC-3 data may be stored as a stand-alone file that contains only the audio data or inside an MPEG or other similar containers (e.g. .m4a).

Standard: Annex E of ATSC A/52:2018 published by the Advanced Television Systems Committee.

Popularity: Common
Support: Good
Type: Lossy
Stand-alone: .ac35
Genre: Home Theater

Dolby Digital Plus (Enhanced AC-3)

Building on the AC-3 platform, Enhanced AC-3 (E-AC-3) - more widely known as Dolby Digital Plus (DD+) - was released in 200639, 15 years after the advent of AC-3.

Dolby Digital Plus is a marked improvement over the AC-3 codec. Supporting an impressive 15 independent audio channels at sampling rates more than 10-times greater than AC-3, it continues to be an excellent choice for many multimedia, multi-channel applications.

No operating system natively recognizes a file extension of .eac3. If you encounter such files, rename the file extension to .ac3

Standard: Annex E of ATSC A/52:2018 published by the Advanced Television Systems Committee.

Popularity: Uncommon
Support: Moderate
Type: Lossless
Stand-alone: .mlp .thd
Genre: Home Theater

Dolby Digital TrueHD (THD/MLP)

Dolby Digital TrueHD is Dolby Laboratories' implementation of Meridian Lossless Packing (MLP) by Meridian Audio. MLP is a lossless successor to AC-3 that debuted in 2008 (just two years after Dolby Digital Plus hit the market), MLP is the standard compression method for DVD-Audio content and Blu-ray Audio discs (~2:1 compression compared to PCM). It is sometimes referred to as "Packed PCM."

MLP/DD TrueHD supports up to eight (8) simultaneous full-bandwidth channels plus 128 audio object types used to generate Dolby Atmos audio content.32 No operating system natively supports MLP or Dolby TrueHD, but many third-party applications exist that incorporate corresponding Dolby codecs, which are only available as licensed products.

Standard: Dolby TrueHD (MLP) bit streams within the ISO base media file format

Files with DTS or THD codec extensions (.dts, .thd) could have been created with any compatible codec.

Popularity: Uncommon
Support: Poor
Type: Lossy
Stand-alone: .dts
Genre: Home Theater

DTS (Digital Theater Systems)

DTS supports seven (7) simultaneous full-bandwidth channels and one (1) low frequency channel. No operating system natively supports it, though numerous third-party applications exist that incorporate the DTS codec. DTS is not open source. It and its variants are owned by DTS, Inc. (formerly Digital Theater Systems, Inc.) and are available only as licensed products.

Standard: Implementation of DTS Audio in Media Files Based on ISO IEC 14496 (2014).

Popularity: Uncommon
Support: Poor
Type: Lossless
Stand-alone: .dts
Genre: Home Theater

DTS-HD Master Audio

Digital Theater Systems High Definition Master Audio (DTS-HD MA) - now that's a mouthful for a codec name - is DTS' equivalent of Dolby's TrueHD. DTS-HD Master is a lossless, proprietary audio codec. It supports seven (7) simultaneous full-bandwidth channels and one (1) low frequency channel. No operating system is natively compatible with it, although some third-party applications incorporate the DTS-HD MA codec. DTS-HD MA is not open source. It is available only as a licensed product.

DTS-HD Master Audio requires substantial bandwidth, in part because it transmits two (2) streams simultaneously: a lossy stream (referred to by DTS as the "core" stream) and an accompanying layer DTS refers to as the "residual" stream, which contains the difference between the original signal and lossy core stream.

Standard: DTS-HD and DTS-HD Master Audio Enabling HD Audio Across an Evolving Media Delivery Landscape (2012).

Popularity: Very Rare
Support: Poor
Type: Lossless
Stand-alone: .dts
Genre: Home Theater

DTS:X

DTS:X is Digital Theater Systems' response to Dolby's Atmos. Like Atmos, DTS:X is a 3-Dimensional audio format that introduces the concept of audio objects, while maintaining backward compatability with traditional channel-based audio encoding methods. DTS:X is not actually a codec; rather it is an extension to the DTS-HD Master Audio codec (it sits on top of that codec). DTS:X streams are backwards-compatible with DTS-HD MA decoders. The latter will ignore the 3D/height object encoding present in DTS:X streams.

Popularity: Very Rare
Support: Poor
Type: Lossless
Stand-alone: .dts
Genre: Home Theater

DTS:X Pro

Currently the pinnacle of home theater encoders from DTS (the company), DTS:X Pro is virtually unknown to consumers. DTS:X Pro provides 30 independent, full range channels, 2 LFE (subwoofer) channels, and audio object-based support (think DTS:X).

DTS:X Pro is arguably the most extreme level of audio differentiation available for the home theater market as of this writing.

For more information on Dolby and DTS codecs, check out Epic Battle of Home Theater Audio: Dolby, DTS, and Auro

Popularity: Popular
Support: Good
Type: Lossless
Stand-alone: .flac
Genre: Music

FLAC (Free Lossless Audio Codec)

Free Lossless Audio Codec or FLAC as its commonly known is the most popular compressed lossless digital audio codec (WAV is the most popular uncompressed lossless file format).40 FLAC offers good compression ratios (across all music types) when compared with other compressed, multi-channel, lossless codecs.

Aside from data compression, FLAC is often compared and contrasted with other lossless formats such as ALAC, Monkey's Audio (APE), WAV, and WMA (lossless version) to some extent. Most advanced users find FLAC has several distinct advantages over these other formats. Notably, it supports both structured and unstructured (arbitrary) metadata (Vorbis Comment)41 and cue sheets.42 Like ALAC, it supports up to eight (8) simultaneous audio channels.

Audiophile forums generally regard FLAC as the de-facto compressed file standard for archival purposes. Why is FLAC regarded by audiophiles as one of the best audio codecs? It offers several advantages over MP3 (the most popular codec):43

  • Lossless
  • Free (open source and non-proprietary)
  • Unencumbered by patents
  • The most widely supported lossless codec (native operating system support)
  • Has an API (helpful for software developers)
  • Supports useful metadata
  • Strong documentation support
  • Streaming capable
  • Multi-channel

FLAC is brought to you by the Xiph.Org Foundation; the same organization who created the Ogg container type, Vorbis codec, and Vorbis Comment metadata container.

Standard: FLAC documentation published by the Xiph.Org Foundation.

Popularity: Very Popular
Support: Very Good
Type: Lossy
Stand-alone: .mp3
Genre: Music

MP3 (MPEG-2 Audio Layer III)

MP3 (officially, MPEG-2 Audio Layer III) is the world's most ubiquitous compressed file format. Technically a lossy audio codec, MP3 is one of a small number of codecs able to stand on its own, yet also frequently found inside various audio container types.

More details on the nuances of MP3 - such as its history and compatibility relative to other MPEG codecs and file formats - may be found in the related article, MPEG Audio Standards. Despite its replacement by the superior AAC standard in 1997, MP3 remained incredibly popular. To this day, it continues to be the dominant audio file format by a massive margin, 25 years after its debut.

An estimated 70% of all online audio files are MP3s.40

Standard: ISO/IEC 11172-3:1993 (aka MPEG-1 Audio Layer III) Information technology -- Coding of moving pictures and associated audio for digital storage media at up to about 1,5 Mbit/s (legacy); ISO/IEC 13818-3:1998 (aka MPEG-2 Audio Layer III) Information technology — Generic coding of moving pictures and associated audio information — Part 3: Audio (current)

There are two (2) generations of MP3. The information above pertains to the current and 2nd generation. The first generation of MP3 was based on the MPEG-1 Layer III audio standard. If you wish to learn more about it, the standards documents mentioned under MP2 and MP1 also pertain to Gen1 MP3.

Most MP3 players can also read the lesser known, previous MPEG Audio format MP2 (MPEG-1 Audio Layer II).

Popularity: Uncommon
Support: Obsolete
Type: Lossy
Stand-alone: .mpc
Genre: Music

Musepack (MPC)

Musepack is an open source, lossy, compressed audio codec based heavily on the MP2 standard (MPEG-1 Layer 2; ISO/IEC 11172-3:1993).

Musepack is mentioned in this document for the sake of completeness. However, its use is discouraged. In the mid-2000's, Musepack was one of the best lossless codecs, however today the most recent version (SV8) is 11 years old (circa 2009) and does not appear to have been maintained by the developer since 2011.44

Standard: Musepack SV8 [2011]

Popularity: Uncommon
Support: Moderate
Type: Lossless
Stand-alone: .opus
Genre: Music

Opus

Opus - also known as Ogg Opus - is an amalgamation of codecs. An early attempt to create a universal codec suitable for transmitting all audio types across the Internet (e.g. speech and music). While often praised for its versatility (Opus supports up to 255 simultaneous audio channels) and the fact it is open source, the end result of Opus' evolution is a bloated and resource intensive product poorly suited to low-end CPU devices.

Opus is often referred to as Ogg Opus, though the term is a misnomer. Ogg Opus and Opus are not synonymous. Opus is a codec and Ogg is a container type. Ogg is not limited to containing only Opus content, but Opus must reside inside a container. Used correctly, Ogg Opus refers to the combination of Opus content inside an Ogg container. Ogg Opus files should have a file extension of .opus, though other file extensions may be used (such as .ogg). Opus content may also be found in Matroska, MPEG-4, WebM, or MPEG-TS containers.

Opus Standards

Opus is huge and complex. Illustrating this point, the Opus specifications span four (4) RFC documents:

Popularity: Low
Support: Poor
Type: Lossless
Stand-alone: .tta
Genre: Music

True Audio (TTA)

True Audio is a lossless, compressed audio codec specifically designed to compress WAV files only. It doesn't do anything else. Recall that WAV is actually a container type that holds uncompressed PCM.

True Audio is free and open source, and sports a high compression rate average of 60% and relatively fast encoding times.45 However, native operating system support is extremely limited (to the point it's almost not worth mentioning). In fact, technically there isn't any as support in Windows requires pre-installation of a codec pack that includes TTA. On-the-other-hand, if you have a Windows-based device with a large collection of WAV files and/or you wish to retain a good sized library of PCM/LPCM content in its original auditory glory while compressing it to conserve storage space, True Audio is certainly worth a look.

Popularity: Uncommon
Support: Limited
Type: Lossless
Stand-alone: No
Genre: Music

Vorbis

Vorbis is a lossless audio codec with built-in support for optional metadata. Vorbis content must reside inside a container and is supported by Ogg and Matroska container types. Ogg Vorbis refers to the combination of Vorbis content inside an Ogg container, and is the most common scenario.

More-or-less the opposite of ALAC when it comes to operating system support, Vorbis is a complex topic. There are several different products with the name "Vorbis" in them, and their similar names frequently lead to confusion among end users. Vorbis, Vorbis Comment, and Ogg Vorbis are each different and independent products.

For a comprehensive explanation of Vorbis, see Vorbis, Vorbis Comment, and Ogg Vorbis Explained

Standard: Vorbis I Specification (current)

Popularity: Uncommon
Support: Poor
Type: Lossless
Stand-alone: .wv
Genre: Music

WavPack

WavPack is an open-source, lossless audio codec designed to compress WAV files (making them smaller) while retaining lossless audio quality.

WavPack is also capable of operating in a "hybrid" mode, where a secondary file is created. In this mode, the primary file is a lossy, compressed copy of the original source. The secondary file acts as a correction file. During playback, the WavPack decoder uses the data in this correction file to restore the details sacrificed during the lossy file's creation. DTS-HD Master Audio and SLS are also capable of similar hybrid operations, though they incorporate both lossy and lossless versions of the original art into a single file.

Popularity: Uncommon
Support: Poor
Type: Lossy
Stand-alone: .wma
Genre: Music

Windows Media Audio (WMA)

Windows Media Audio (WMA) is a blanket term referring to a family of proprietary audio codecs developed by Microsoft Corporation.

This section describes WMA, a compressed, lossy audio format, and the most common of the WMA codecs. Like many things created by Microsoft, WMA files are a bit of an oddball. First, there are four (4) WMA audio codecs: WMA, WMA Pro, WMA Lossless, and WMA Voice. These codecs must reside inside a container.

The WMA Voice codec is not discussed in this document.

WMA (lossy) files have very limited native operating system support in Windows and Linux. A WMA file is nearly always incorporated into an Advanced Systems Format (ASF) container, a proprietary Microsoft container format for digital audio and/or digital video. The file's extension will be .wma, .wm, or .asf.

A .wma file extension is indicative of Windows Media Audio content inside an ASF container.

Popularity: Uncommon
Support: Poor
Type: Lossless
Stand-alone: .wma
Genre: Music

Windows Media Audio - Lossless (WMA Lossless)

Windows Media Audio (WMA) is a blanket term referring to a family of proprietary audio codecs developed by Microsoft Corporation.

This section describes WMA Lossless, a compressed, lossless audio format. It must reside inside a container.

WMA (lossless) files are natively supported in Windows operating systems only. A WMA file is nearly always incorporated into an Advanced Systems Format (ASF) container, a proprietary Microsoft container format for digital audio and/or digital video. The file's extension will be .wma, .wm, or .asf.

Popularity: Very Rare
Support: Very Poor
Type: Lossy
Stand-alone: .wma
Genre: Music

Windows Media Audio Professional (WMA Pro)

Windows Media Audio (WMA) is a blanket term referring to a family of proprietary audio codecs developed by Microsoft Corporation. Like many things created by Microsoft, WMA files are a bit of an oddball. First, there are four (4) WMA audio codecs: WMA, WMA Pro, WMA Lossless, and WMA Voice. These codecs must reside inside a container.

This section describes WMA Pro, a compressed, lossy audio format released in 2008, with minimal adoption due to a lack of widespread hardware support and the fact other codecs are more readily available that accomplish similar goals. WMA Pro is an 8-channel (7.1) capable codec with strong audio sampling rates (24-bit words at up to 96 Khz sampling rate).

WMA Pro files are natively supported in Windows operating systems only. A WMA file is nearly always incorporated into an Advanced Systems Format (ASF) container, a proprietary Microsoft container format for digital audio and/or digital video. WMA Pro content is normally contained inside a PIFF container (.piff file extension). "PIFF" stands for Protected Interoperable File Format, a multi-media container type.

Low Bandwidth Codecs

These codecs are designed for circumstances where bandwidth is limited and/or network connections are likely to experience high latency. They are well suited to inconsistent/unreliable, but constant or near-constant network connection requirements such as streaming media over mobile phone networks and industrial environments.

Popularity: Common
Support: Limited
Type: Lossy
Stand-alone: No
Genre: Mobile

AAC+ (Advanced Audio Coding Plus/HE-AAC)

Advanced Audio Coding Plus (AAC+), also known as High-Efficiency Advanced Audio Coding (HE-AAC), is a more highly compressed version of AAC. Based on their names, you would think AAC+ would be an improved version of AAC. Whether or not that's the case depends on one's perspective. AAC+ was designed for low-bitrate streaming applications, such as playing music over a mobile phone network. Its intent is to provide the perception of "good" audio quality on devices that are far from optimal when it comes to audio fidelity. AAC+ is most likely to be found in 3GPP, ADTS, and MPEG-4 containers.

AAC Plus is designed to deliver comparable audio quality (compared to AAC) via smaller data bursts. This makes it conducive to high-latency environments such as ground-to-ground, ground-to-satellite, and satellite-to-air wireless networks where at least one network peer is in motion. Since the data is highly compressed and the file sizes are small, AAC+ is generally a superior method of audio transmission for unstable peer connections. And because typical target devices have very limited audio capabilities to begin with, any reduction in audio fidelity is nearly undetectable to the end user. Examples of industries where AAC+ is widely used include mobile cellular phone networks, Internet radio stations, and aircraft-to-satellite-to-ground station peer networks. AAC+ supports up to 48 simultaneous full range channels and 16 LFE46 channels (just like AAC).

AAC+ is NOT a "better" version of AAC when it comes to audio quality. Quite the opposite.

A primary strength of AAC+ (HE-AAC) is its focus on network connection stability. This makes AAC+ adept at navigating unstable network connections, such as those found with mobile devices which tend to be in motion.

Standard: MPEG-4 Part 3: Audio — Amendment 1: Bandwidth extension (ISO/IEC 14496-3:2001)

Popularity: Uncommon
Support: Poor
Type: Lossless
Stand-alone: .als
Genre: Over-the-Air

ALS (Audio Lossless Standard)

Audio Lossless Coding Standard - also known as Audio Lossless Standard or Low Overhead Audio Transport Multiplex (LATM) - is a lossless, single, non-multiplexed AAC stream encoding technique modeled after AAC. In many respects, ALS resembles AAC+.

ALS is treated as a codec, but is actually an extension to the AAC audio codec defined under the MPEG-4 (Part 3) standard in ISO/IEC 14496-3:2005/Amd 3:2006 [2006]. It is mentioned here as if it were an independent codec because it is possible to encounter files encoded with ALS and bearing an .als file extension. ALS files are backwards-compatible with AAC.

ALS is not directly related to SLS (Scalable Lossless Coding Standard), although they are both MPEG-4 Part 3 audio codec extensions. ALS was developed by MPEG while SLS was a joint 3rd party initiative later adopted by MPEG and incorporated into the MPEG-4 Part 3 set of standards.47 The ALS decoder cannot interpret SLS content. However, the ALS decoder is backward compatible with AAC content.

ALS is backwards-compatible with both MPEG-4 and MPEG-2 variants of AAC.

ALS is designed specifically for broadcasting high-quality streams of audio data over low-bandwidth connections, such as over-the-air digital television (HDTV) signals. This is an important distinction from an implementation viewpoint, compared to the use-case scenarios of AAC+. They are similar, but solve different problems.

ALS prioritizes signal quality over connection robustness, making it better suited to stationary terrestrial broadcasters and receivers when compared with higher quality audio codecs such as AAC. Mobile connections are less likely to be concerned with factors such as signal quality/depth due to the fact a typical mobile form factor is unsophisticated from an audio/visual perspective (e.g. small screens). Thus, less data is required to reproduce acceptable results (to the end user) on the receiving device. The markets where ALS is most likely to be a better fit also tend to have receivers employing much more sophisticated audio/visual reproduction capabilities (e.g. large screen TVs), making signal quality and depth a higher priority as the end user is far more likely to perceive artifacts created by data loss.

Standard: Original MPEG-4 specification ISO/IEC 14496-3:2005; current specification ISO/IEC 14496-3:2019.

Popularity: Uncommon
Support: Limited
Type: Lossy
Stand-alone: .amr
Genre: Mobile

AMR (Adaptive Multi-Rate)

AMR is an abbreviation for Adaptive Multi-Rate Audio Codec. A proprietary audio codec optimized for speech.

AMR content may be found inside 3GP containers, where is not uncommon to find it used to store speech alongside raw video data in a separate sub-container or video codec in the same file. AMR is rarely encountered outside a container, as a stand-alone file with AMR content. However, when that is the case, its default file extension is .amr.

AMR is supported by a number of third-party media players, including VLC Player, foobar2000, QuickTime, RealPlayer. It is natively supported by iOS, Android, macOS, and ffmpeg (a Linux media player).

Standard: RFC 4867: RTP Payload Format and File Storage Format for the Adaptive Multi-Rate (AMR) and Adaptive Multi-Rate Wideband (AMR-WB) Audio Codecs (2007).

Popularity: Rare
Support: Good
Type: Lossy
Stand-alone: .mp1
Genre: Music

MP1 (MPEG-1 Audio Layer I)

MP1 (officially, MPEG-1 Audio Layer I) is a predecessor to MP2. MP1 is limited to 2-channel stereo and should be avoided. MP3, AAC, or AAC+ (HE-AAC) should be used instead.

Standard: (aka MPEG-1 Audio Layer I) Information technology -- Coding of moving pictures and associated audio for digital storage media at up to about 1,5 Mbit/s (legacy); ISO/IEC 11172-3:1993 (Information technology — Coding of moving pictures and associated audio for digital storage media at up to about 1,5 Mbit/s — Part 3: Audio).

Popularity: Rare
Support: Good
Type: Lossy
Stand-alone: .mp2
Genre: Music

MP2 (MPEG-1 Audio Layer II)

MP2 (officially, MPEG-1 Audio Layer II) is a predecessor to MP3. MP2 is limited to 2-channel stereo and should be avoided. MP3 or AAC should be used instead.

Standard: MPEG-1 Audio Layer II in Information technology -- Coding of moving pictures and associated audio for digital storage media at up to about 1,5 Mbit/s.

Popularity: Uncommon
Support: Poor
Type: Lossless
Stand-alone: .sls
Genre: Over-the-Air

SLS (Scalable Lossless Standard)

Scalable Lossless Coding Standard (SLS), also known as Scalable Lossless Standard and HD-AAC (High Definition Advanced Audio Coding).48

SLS is similar to ALS; with a few twists that differentiate it significantly. SLS and ALS are extensions to the MPEG-4 version of AAC. While ALS is an MPEG standard.47

SLS is an extension of the MPEG-4 Audio standard (even though SLS has the word "standard" in its name).

SLS has two (2) operating modes:

  1. Lossy layer (AAC) + lossless correction layer49
  2. Lossless layer only

The standard (first) method contains both "core" (lossy) and "non-core" (lossless adjunct) layers in the same file or stream, resulting in a smaller file that is backward compatible with AAC based decoders such as ALS, AAC (MPEG-3: Part 3), and AAC-LC (MPEG-2: Part 7) decoders while retaining lossless audio quality (available to SLS decoders only). This concept is identical to the technique deployed by the DTS-HD Master Audio codec (DTS-HD MA refers to its supplemental stream as the "residual" layer).

When a SLS stream is encoded in lossless (only) mode, a SLS decoder is required to read the file.50

SLS files of either type are considerably larger than ALS or AAC files due to their lossless correction layer.

Why would anyone want to use the lossless-only encoding method in SLS? The only compelling reason to operate the SLS codec in lossless-only mode is speed. Both encoding and decoding processes are faster, at the expense of a larger file size and eliminating backward compatibility with older AAC-based decoders.

Standard: Original MPEG-4 Part 3 specification ISO/IEC 14496-3:2005/Amd 3:2006; current specification ISO/IEC 14496-3:2019.

Audio Containers

A multimedia container is a digital object that stores one or more audio and/or video streams encoded via one or more codecs. Audio containers are containers composed of one or more codec data streams. Although it is not required, most audio containers also encompass one or more distinct sets of metadata. Aside from metadata, containers do require fields that describe their contents.

* User-defined arbitrary/contextual metadata, not required metadata.
Audio File Containers
Container File
Extensions
Supported
Codecs
Meta
data*
Native Support by Operating System
Win Lin Mac iOS And
3rd Generation Partnership Project (3GPP) .3gp | .3gpp AAC | AAC+ | AMR No Yes No No Yes Yes
Audio Data Interchange Format (ADIF) .adts AAC No Yes No Yes No Yes
Audio Data Transport Stream (ADTS) .adts AAC No Yes No Yes No Yes
Audio Interchange File Format (AIFF) .aif | .aiff
.aifc
PCM29 Yes Yes No Yes No No
Advanced Systems Format (ASF) .asf
.wma
.wm
WMA | DTS | FLAC
AMR | MP1 | MP2
MP3 | AC3 | AAC
Opus | FLAC | ALAC
LPCM | ADPCM | μ-law
Yes Yes No No No No
Audio (Au) .au | .snd ADPCM
LPCM | μ-law51
No Yes No Yes No No
Apple Core Audio Format (CAF) .caf AAC | ALAC
FLAC25,26 | LPCM
MP3 | Opus9
ADPCM
No No No Yes Yes No
Exchangeable Image File Format (EXIF) .wav PCM | ADPCM
LPCM | μ-Law51
Yes Yes No Yes No No
Material eXchange Format (MXF) .mxf AC-3 | MP1 | MP2
MP3 | AAC | PCM
LPCM
Yes No No No No No
Matroska .mka MP3 | AAC | AC3
THD | DTS | OGG
MPC | PCM | WavPack
FLAC52 | Vorbis
TTA | Opus53 | MP2
Yes Yes Yes No No Yes2
MPEG-4 Audio (MP4) .m4a
.alac
AAC | AAC+ | ALS
MP3 | ALAC | MP2
SLS | E-AC-3 | DTS
Opus
No Yes No Yes Yes No
Ogg .ogg
.oga | .ogga
FLAC54 | Opus
Vorbis
Yes Yes Yes No No Yes2
Resource Interchange File Format (RIFF) .riff PCM | WAV Yes Yes Yes Yes No No
WAVe (WAV)55 .wav | .wave56 PCM29 | ADPCM | LPCM Yes Yes Yes No No Yes
WebM .webm Opus | Vorbis Yes Yes8 No No No Yes

3GPP (3rd Generation Partnership Project)

3rd Generation Partnership Project (3GPP) is a telecommunications industry consortium and standards development organization created for the express purpose of coordinating mobile wireless broadcast technologies, such as 5G cellular radio networks. Among the group's accomplishments is the adoption of a multimedia container format designed specifically to address the needs of mobile networks. It is predominantly used with cellular devices to transmit audio and video content. 3GPP files should always bear a file extension of .3gp.57 Some legacy files may have a .3gpp extension, especially on macOS.

3GPP audio must conform to a MPEG-2 or MPEG-4 AAC-based audio codec. While this can be AAC, it will normally be a deriviative of AAC due to 3GPP's focus on low-bandwidth data transfer solutions. Compatible codecs include:

  • AAC
  • eAAC+ (Enhanced AAC Plus)
  • AMR
  • AMR-WB (Adaptive Multi-Rate Wideband) for speech
  • AMR-WB+ (Extended Adaptive Multi-Rate Wideband) for speech and/or stereo music; allows higher sampling rates than AMR or AMR-WB
  • TwinVQ; very low bandwidth speech codec

Standards:

  • Primary: ETSI TS 126 244 V14.1.0 Release 14; Digital cellular telecommunications system (Phase 2+) (GSM); Universal Mobile Telecommunications System (UMTS); LTE; Transparent end-to-end packet switched streaming service (PSS); 3GPP file format (3GP) (3GPP TS 26.244 version 14.1.0 Release 14) [2018]
  • RFC 3839: MIME Type Registrations for 3rd Generation Partnership Project (3GPP) Multimedia files (2004)
  • RFC 6381: The 'Codecs' and 'Profiles' Parameters for "Bucket" Media Types (2011)
  • ISO/IEC 14496-12:2015 (Information technology — Coding of audiovisual objects — Part 12: ISO base media file format) [2015]
File extension: .3gp
Contextual metadata (structured): No
Contextual metadata (unstructured): No

Audio Data Interchange Format (ADIF)

Audio Data Interchange Format (ADIF) is a limited-use multimedia container restricted to AAC content only. Intended as a storage and local-use media bucket, ADIF falls under the MPEG-2 Part 3 AAC standard (ISO/IEC 13818-3:1998). Like its cousin ADTS, ADIF is also referenced in the MPEG-4 Audio standard (ISO/IEC 14496-3), though on an informational basis only (though generally speaking ADIF will work with AAC-LC and HE-AAC decoders). Making matters a bit more confusing, ADIF container files bear the .aac file extension. ADIF is the oldest file formats belonging to the AAC family of codecs.

Standards:

  • Current Specifications:
    • ISO/IEC 14496-3:2009 Information technology — Coding of audio-visual objects — Part 3: Audio. (1 September 2009). Moving Picture Experts Group (MPEG). 4th Edition. Section 1.A.3.1: Audio_Data_Interchange_Format (ADIF).
  • Original Specifications:
    • MPEG-2: ISO/IEC 13818-7:2003. Information technology -- Generic coding of moving pictures and associated audio information -- Part 7: Advanced Audio Coding (AAC).
File extension: .aac
Contextual metadata (structured): No
Contextual metadata (unstructured): No

Audio Data Transport Stream (ADTS)

Audio Data Transport Stream (ADTS) is a rare, limited-use multimedia container restricted to AAC content only.58 Intended as a transmission (streaming) bucket, ADTS falls under the MPEG-2 Part 7 AAC standard (ISO/IEC 13818-7:2006). Though also referenced in the MPEG-4 Audio standard (ISO/IEC 14496-3), this is on an informational basis only (though generally speaking ADTS works with AAC-LC and HE-AAC). Making matters a bit more confusing, ADTS files with only audio content may bear the file extension .aac or .m4a. ADTS files containing both audio and video content must use the extension .mp4.

ADTS is one of the oldest file formats belonging to the AAC family of codecs. The use of the .adts file extension is currently discouraged. If at all possible, .aac should be used instead. Programs capable of reading AAC files should correctly interpret an ADTS file, since the underlying audio content must be AAC and the ADTS container type is very simple, however such support is not guaranteed. Introduced in the MPEG-2 Part 7 standard, support for ADTS is not required under the MPEG-4 standards that subsequently took over the AAC protocols. Thus, the subject of modern ADTS file support is murky. Given the passage of time, it is conceivable you could encounter arcane ADTS files which - even though they are technically an AAC file format - a modern player may not be capable of interpreting their content if the decoder is MPEG-4 compliant only.35 This is an unlikely scenario, but is plausible.

AAC-LC stands for "Advanced Audio Coding - Low Complexity," where "low complexity" means the standard version of AAC (version 1 defined under MPEG-2 Part 7, and/or version 2 defined under MPEG-4 Part 3).

Standards:

  • Current Specifications:
    • ISO/IEC 14496-3:2001 Information technology — Coding of audio-visual objects — Part 3: Audio. (15 December 2001). Moving Picture Experts Group (MPEG). 2nd Edition. Section 1.A.3.2: Audio_Data_Transport_Stream (ADTS).
  • Original Specifications:
    • MPEG-2: ISO/IEC 13818-7:2003. Information technology -- Generic coding of moving pictures and associated audio information -- Part 7: Advanced Audio Coding (AAC).
File extension: .adts
Contextual metadata (structured): No
Contextual metadata (unstructured): No

AIFF (and AIFC)

Audio Interchange File Format (AIFF) is an uncompressed audio file container format, usually in the form of Pulse Code Modulation (PCM) or Linear Pulse Code Modulation (LPCM). AIFF was invented by Apple, Inc. in 1988. Microsoft developed a very similar, competing container format the same year called, Resource Interchange File Format (RIFF). Both AIFF and RIFF trace their origins to an ancient multimedia container standard introduced by Electronic Arts in 1985, called IFF (Interchange File Format). Designed to facilitate information sharing between incompatible systems, IFF was one of the very first file standards to incorporate metadata.

AIFF has an upper file size limit of 4 GiB.

Today, support for this standard has waned, though in spite of its age AIFF remains the only comparable competitor to Microsoft's WAVE file format (which itself is based on RIFF). While still popular with musicians, outside that user group AIFF is rarely found in the wild.

Standard: Camera and Imaging Products Association (CIPA). "Exchangeable image file format for digital still cameras: Exif Version 2.32." (May 2019). CIPA DC-008-2019. Section 5: Exif Audio File Specification.

File extensions: .aif | .aiff | .aifc59
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes50

.aifc files are compressed AIFF files.

Advanced Systems Format (ASF)

Advanced Systems Format (ASF) is a multimedia container format developed by Microsoft. Although based upon RIFF, ASF was designed to replace RIFF as its successor.51

Microsoft's original intent for ASF was to develop a multimedia container for world-class audio/visual streaming that it owned and could successfully license (and generate income off the associated fees). In fact, its name was originally, "Advanced Streaming Format," but Microsoft changed the name for some reason. The name change has actually generated a lot of confusion across the Internet, with many sources using the incorrect name. This issue is discussed in greater depth in the related article, Video Codecs, Containers, and File Extensions.

If you are on a Windows-based operating system, ASF's support for audio codecs is unparalleled. It supports more audio codecs than any other container (even Matroska). ASF also has arguably the most flexible metadata capabilities (although they are often overlooked as its native structured contextual metadata is quite weak). On non-Windows based operating systems, you will find better support from Matroska and/or MPEG-4 containers.

Standard: Advanced Systems Format (ASF) Specification v.1.20.03 [2004] (current specification)

File extensions: .asf | .wma
Contextual metadata (structured): Yes
Contextual metadata (unstructured): No

Au (Audio)

Au - short for Audio - is an audio container format with very limited codec support. Invented by Sun Microsystems, Au is still used on UNIX systems. Although primitive by modern digital audio standards, Au's usefulness has actually grown over time as the codec now supports LPCM as a lossless encoding format and several lossy codecs, including ADPCM and μ-law.52

Au is represented via the file extensions .au or .snd (SouND). Outside of UNIX based platforms it is relatively useless, though it enjoys minimal support in Microsoft Windows and macOS (ADPCM). Au is a PCM container and supports a maximum of 2 channels.

Standard: Camera and Imaging Products Association (CIPA). "Exchangeable image file format for digital still cameras: Exif Version 2.32." (May 2019). CIPA DC-008-2019. Section 5: Exif Audio File Specification.

File extensions:
Contextual metadata (structured): No
Contextual metadata (unstructured): No

Broadcast WAVE Format (BWF)

Broadcast WAVE Format (BWF) is an audio media file standard derived from RIFF/WAVE.

Originally developed by the European Broadcast Union (EBU), BWF is and always has been intended as a long-term storage file format for the purpose of archiving and transferring broadcast media. Like WAVE, it has a 32-bit architecture, limiting file size to less than 4 GB.

Although backward compatible with WAVE, full support in media players is not guaranteed. At its core, BWF files are WAVE files with an additional chunk in the file called the Broadcast Audio Extension Chunk, which performs two functions: it identifies the file as a BWF formatted file; and it contains skeleton metadata specified in the BWF specification. Otherwise, BWF and WAVE files containing PCM or LPCM audio data are synonymous.

So, what's the big deal with BWF if it's just a re-branded WAVE file? BWF is more than just a WAVE file with a different file extension. BWF makes it possible to incorporate other (non-PCM) audio codecs into a WAVE/RIFF-type container, and it has more comprehensive technical metadata.

Like any other RIFF standard, BWF files may incorporate independent metadata schemas (e.g. XMP, ID3) inside of custom chunks, but support for such metadata is tenuous and varies by file reader. BWF has very limited structured, contextual metadata as part of its standard:

  • COMMENT field metadata (ASCII)
  • Title (max 256 chars, ASCII)
  • Date (10-digit)
  • Operator (who digitized file; 64-char ASCII)

An advantage of BWF is its built-in compatibility with non-PCM/LPCM/ADPCM sound formats. This is the key that allows certain other audio codec data to be stored inside a BWF file. Which other codecs? Specifically, MPEG formats (MPEG Layer I or II). Alas, BWF only supports a maximum of 2 channels, thus only Layer I and II of MPEG-1 and MPEG-2 are supported (i.e. MP1 and MP2, but not MP3).

Astute observers will note an MPEG-in-a-BWF file entails placing a compressed, lossy audio stream inside an uncompressed, lossless container. Seems a bit pointless, eh? So, what is the point? Recall that BWF was designed as an archival file format. Neither MPEG-1 Audio Layers I nor II support metadata, but BWF does. Therefore, it can make sense to store such files inside a BWF container (particularly for archival purposes).

BWF files are limited to a maximum size of 4 GB.

The first version of BWF was released in 1997 (EBU TECH 3285 Version 0). The current specification is BWF Version 2, published in 2011.

Current specification: BWF - EBU TECH 3285 Version 2 (2011)

BWF version history is confusing. The EBU (European Broadcasting Union) created and owns BWF, while the ITU (International Telecommunication Union) provides technical advice that is incorporated by the EBU most of the time.

  1. EBU TECH 3285 Version 0 (1997)
  2. EBU TECH 3285 Version 1 (2001)
  3. ITU-R BS.1352-3: BWF File Specification Recommendations from ITU regarding MPEG-1 Audio Layer I/II support; Version 4 (2007) [Fourth technical paper from ITU providing guidance to EBU's Supplement 1 (link below)]
  4. EBU TECH 3285 Version 2 (2011) [Current spec]
  5. Seven (7) different EBU TECH 3285 supplements (1997-2018)
File extension: .bwf
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

Broadcast WAVE Format 64 (BW64)

Broadcast WAVE Format 64 (BW64) is the 64-bit companion to BWF and an alternative to RF64.

The preeminent standard in 64-bit RIFF/WAVE container file formats, BW64's architecture allows file sizes to exceed BWF's 4 GB size limit and supports the advanced (2019) version of the ITU's Audio Definition Model (ADM); a technical metadata format that defines multiple loudspeaker positions for replay of Channel-based, Scene-based, and Object-based audio.

Specifications:

  • Current BW64 specification: ITU BS.2088-1 (2019) "Long-form file format for the international exchange of audio programme materials with metadata." [Version 1]
  • Original BW64 specification: ITU Recommendation BS.2088-0 (2015) [Version 0]
File extensions: .wav | .bw64
Contextual metadata (structured): No
Contextual metadata (unstructured): Yes

Apple Core Audio Format (CAF)

Core Audio Format is a flexible audio container format produced by Apple, Inc. Note that while newer versions of Apple's Mac operating system support CAF natively, older versions require QuickTime version 7 or later to be installed first. A potential advantage of the CAF format compared with most other audio containers is its file size is virtually unlimited.

Standard: Apple's CAF specification is only available online, at Apple Core Audio Format Specification 1.0

File extensions: .caf
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

Exchangeable Image File Format (EXIF)

The Exchangeable Image File Format (EXIF) specification was originally designed for adding metadata to photographs. Over time it expanded to include the ability to record audio data. Its primary purpose with regards to audio is to associate audio data with photographs. EXIF may also be used exclusively to record certain types of PCM-based audio data. EXIF is capable of reading WAVE container files and programs capable of understanding WAVE containers are capable of decoding EXIF audio files. EXIF follows the RIFF (Resource Interchange File Format) standard as a foundation for audio files (also known as RIFF/WAVE). A potential advantage for some audio recording circumstances of EXIF is its structured metadata.

EXIF audio files use the .wav file extension and must follow the RIFF/WAVE container format (WAVE) standard. The following audio codecs are supported:

EXIF-based metadata is optional. When present, it is incorporated into RIFF "INFO" chunks, as defined in the EXIF specification (currently version 2.32; May 2019).

Standard: Camera and Imaging Products Association (CIPA). "Exchangeable image file format for digital still cameras: Exif Version 2.32." (May 2019). CIPA DC-008-2019. Section 5: Exif Audio File Specification.

File extensions:
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

Material eXchange Format (MXF)

Material Exchange Format (MXF) is a multimedia container format for video and audio media defined by a set of SMPTE standards. MXF is very rarely found in use by consumers. It was designed to be and is most often found in corporate environments, where it is popular as an archival file type. In fact, MXF was developed specifically to work with the Advanced Authoring Format file format; a professional, cross-platform information exchange protocol designed for post-production video sharing and editing. How are they related? AAF is designed to be a working standard, where a work-of-art is not yet finalized. MXF is intended as an archival format, for the long-term storage of completed work.

MXF is mentioned here for completeness, and because consumers and digital audio enthusiasts may occassionally encounter a MXF file.

Standard: ST 377-1:2011 - SMPTE Standard - Material Exchange Format (MXF) — File Format Specification. (7 June 2011).

File extensions: .mxf
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

Matroska

Matroska is a very popular, open source, multimedia container. It is the basis for WebM. In fact, Matroska is the closest thing to a universal container type you will find. By design, it currently supports virtually all known audio and video compression formats.

Standard: Matroska Specifications

File extensions: .mka
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

MPEG-4 Audio

MPEG-4 Audio - commonly known as MPA or .mpa is part of the MPEG-4 family of digital multimedia container formats.

While the only official filename extension defined by the MPEG-4 standard is .mp4, various filename extensions are commonly used to indicate intended content (such as audio). It is a common practice for an .mpa file extension to specifically denote a file contains content related only to the audio portion of the MPEG-4 standards.

MPEG-4 Audio and 3GPP are closely related. Both are derived from the same ISO/IEC base media file format standard.

Standard: MPEG-4 MP4 Audio/Video Standard ISO/IEC 14496-14:2020 Information technology — Coding of audio-visual objects — Part 14: MP4 file format. [2020]54

File extensions: .m4a | .alac
Contextual metadata (structured): Varies (depends on implementation)55
Contextual metadata (unstructured): No

Ogg

Ogg is a popular, open source, multimedia container format. A native file and stream format for the Xiph.org multimedia codecs (e.g Vorbis and FLAC). Ogg is a stream oriented container. It is designed such that it may be written and read in one pass, making it a natural fit for internet streaming, compared with file-based container formats.

Note that Ogg does not contain contextual metadata, though of course the audio and video codecs stored inside of it can.

Ogg file extensions follow an unofficial, but widely adopted pattern. New users of Ogg should endeavour to follow this routine in order to avoid confusion when archiving and sharing files.

Vorbis encoded audio .ogg
Ogg audio .ogx
Ogg audio on macoS .OggX
Ogg video .ogv
Ogg video on macOS .OggV
Ogg multimedia .ogg | .oga
Ogg multimedia on macOS .OggA
Ogg with speech content .spx

Standards:

  • Primary: RFC 3533: The Ogg Encapsulation Format Version 0 [2003]
  • RFC 8486: Ambisonics in an Ogg Opus Container [2018]

Additional Information: Ogg bitstream overview

File extensions: .ogg | .oga | .ogga
Contextual metadata (structured): No
Contextual metadata (unstructured): No

Ogg is not the same as Ogg Vorbis (the latter is Ogg container + Vorbis codec).

Resource Interchange File Format (RIFF)

Resource Interchange File Format (RIFF) is a generic (audio/video) container format. RIFF is rarely encountered today in its pure form, though it is the basis for AIFF and WAVe container file formats. The continued use of RIFF based files is generally discouraged.

RIFF's origins date back to an ancient multimedia container standard introduced by Electronic Arts in 1985, called IFF (Interchange File Format). Designed to facilitate information sharing between incompatible systems, IFF was one of the very first file standards to incorporate metadata. RIFF derivatives have a maximum size limit of 4 GiB. Very old operating system/hardware platforms may not be capable of processing a RIFF, AIFF, or WAVe file over 2 GiB (up to the 4 GiB hard limit) unless the file's extension is .rf64 or .wav64, denoting a 64-bit addressing method within the file. Generally speaking, modern hardware/software combinations built after 2005 are unlikely to have issues with >2 GiB file sizes, and nearly all systems built in 2009 or later should not have any issues.

Standards:

File extensions: n/a
Contextual metadata (structured): No
Contextual metadata (unstructured): Yes

RIFF File 64 (RF64)

A 64-bit WAVE file format, RF64 functions exactly like a WAVE file. Their difference lies in RF64's 64-bit addressing architecture (compared with WAVE's 32-bits), which allows RF64 files to cross WAVE's 4 GB file size threshold. WAVE's limitation on file size is due to its inability to address internal data pointers in a single file beyond 32-bits, limiting its reach to 4 GB of data.

Official support for RF64 ended in 2018.66

RF64 is a multi-channel capable container created by the European Broadcasting Union (EBU), an audio standards authority focused on terrestrial broadcasting technologies and formats.57 RF64 supports up to 18 surround sound channels, plus a parallel stereo down mix for a combined total of 20 channels of bandwidth.56

RF64 is a 64-bit extension of the RIFF/WAVE standard. It is backward compatible with the 32-bit BWF and WAVE container and file format standards).58

Belonging to a family of WAVE/RIFF based containers indirectly associated with WAVE, the RF64 standard has a convoluted history. The European Broadcast Union (EBU) launched RF64 in 2007 in an effort to address BWF's file size constraints (4 GB hard limit). Two (2) years later in 2009, EBU applied minor tweaks in its first revision. Then in 2015, the International Telecommunication Union (ITU) - a competing multi-national industry standards organization - released a similar audio container and file format called BW64 (short for "BWF-64"). By 2018, the EBU was forced to acknowledge ITU's version held a markedly superior design and it no longer made sense for both organizations to publish such similar, competing standards. That year the EBU issued its final RF64 technical standard (EBU Tech 3360 Version 2 - FINAL), wherein it relinquished and assigned the responsibility of setting broadcast media standards for 64-bit RIFF/WAVE file and container formats henceforth to the ITU. RF64 was effectively dead.

RF64 is backward compatible with WAVE and BWF.

Standards:

  • Original Specification: EBU TECH 3306-2006: RF64 Specification Version 0 (2006)
  • Second Edition: EBU TECH 3306-2007 RF64 Version 1.0 (2007)
  • Latest Specification: EBU TECH 3306-2009 RF64: An extended File Format for Audio - Technical Specification Version 1.1 (2009)
  • Sunset Provision: EBU TECH 3306-2018 Version 2 (2018)
File extensions: .rf64
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

WAVE (WAV)

WAVE (frequently referred to as Wav or .wav) is both a container type and an audio file format type. WAVE files store uncompressed audio using Pulse Code Modulation (PCM) or Linear Pulse Code Modulation (LPCM). They may also store compressed ADPCM codecs.59 Released in 1992 by Microsoft Corporation as the official audio file format of the Windows 3.1 operating system, WAVE remains quite popular in spite of being one of the oldest legacy audio container types still in use today.

WAVE is the iconic implementation of the RIFF standard, and is sometimes referred to as RIFF/WAVE.

Initially limited to 2-channels, in the early 2000's Microsoft partially re-engineered WAVE. The most notable effect was the removal of its restriction on number of channels. However, since it is based on a 32-bit architecture, WAVE does have a hard limit of 65,536 channels by virtue of its inability to index beyond that number.

WAVE's popularity as a container stems not only from its venerable position historically, but also due to its ability to incorporate custom "chunks" of data with virtually whatever content you'd like. Thus, for instance XMP, ID3, or Vorbis Comment metadata may be included in any given WAVE file. Likewise, one could conceivably include non-PCM based audio content, such as MP3 content, for example. Other codecs are theoretically possible. If custom metadata is included in a WAVE file chunk, it is considered unstructured regardless of how it is presented.

The practice of incorporating non-standard data into a WAVE file is discouraged. Such content may be unplayable by other applications. Caution is advised if you plan on distributing WAVE formatted content to a wide audience.

WAVe files have a maximum size of 4 GB.

Standard: RIFF documentation (as amended) serves as the official WAVE standard.

See also:

File extensions: .wav | .wave
Contextual metadata (structured): Yes
Contextual metadata (unstructured): Yes

WebM

WebM is a lossy multimedia container format stemming from an open-source project of the same name (the WebM Project), founded by Google in 2010. WebM is a subset of the Matroska container format, which is also open-source. WebM containers may store audio, video, or a combination of audio and video.

The audio codecs supported by WebM are Opus and Vorbis.

Standard: WebM documentation website

File extensions: .webm
Contextual metadata (structured): Yes
Contextual metadata (unstructured): No

Honorable Mentions

A few codecs, containers, and file types worth mentioning for the sake of completeness. Some of these are not really audio file types at all, but are often mis-understood by consumers as being some sort of audio file type.

Uncommon Audio Codecs and Containers

These file types may be encountered from time to time. If you are not already familiar with them, you are better off avoiding them.

Real Audio (Real Networks)

Real Audio is an audio codec developed by Real Networks, Inc. It is a proprietary codec format requiring the licensing of the company's decoder in order to play back Real Audio content. With the rise of open-source audio codecs and containers beginning in the late 1990's, Real Audio's popularity faded rapidly. Today, it is very rarely encountered. There are no appreciable gains to encoding audio data in the Real Audio format as a variety of open-source encoders offer equal or better compression and fidelity, coupled with much wider and royalty-free codec support.

Obsolete Codecs

Audio codecs mentioned below are considered obsolete. If you have content created with one of these codecs, it is recommended you determine whether or not you are interested in retaining the content. If so, attempt to locate a conversion program and convert the content into a more widely supported audio codec of your choice.

Shorten (.shn)

Shorten is a lossless audio codec and file type released in 1993 and last updated in 2007. From an audio standpoint, it is obsolete compared with more modern and ubiquitious codecs with equal or better performance, such as FLAC and Monkey's Audio (APE). Shorten is not supported by most decoders, though several programs permit converting Shorten files to modern alternatives such as FLAC.50

Speex (.spx)

.spx files are created with the Speex audio codec, which is designed explicitly for speech recordings. This codec is now considered obsolete, having been replaced by Opus.51

Obsolete Containers

Audio containers mentioned below are considered obsolete. If you have content created with one of these containers, it is recommended you determine whether or not you are interested in retaining the content. If so, attempt to locate a conversion program and convert the content into a more widely supported audio container of your choosing.

XMF (.xmf)

XMF is an abbreviation for eXtensible Music Format, an audio container specification developed by the MIDI Manufacturers Association and closely related to the MIDI standard. XMF is a chunk-style container format.

Released in 2001, XMF files may contain one or more specific file types. The XMF file serves as a wrapper for certain content and file types defined in the XMF Specification. XMF files are limited to standard MIDI files, DownLoadable Sound (DLS) instrument files, WAV files, and/or combinations of the aforementioned file types. The purpose of XMF files is to create a collection of all resources needed to present a musical piece, from MIDI (metadata/descriptive) files, DLS instrument files, and audio content preserved in a RIFF/WAVE container (embedded inside the XMF container).

Due to the passage of time and waning interest, XMF files should be considered unsupported from an encoding standpoint. No modern operating system natively supports XMF files, and even third-party applications that support it are very rare (if you can find any at all). XMF currently only supports the original MIDI standard (i.e. not MIDI version 2.0).

More information about XMF may be found on the MIDI Association's website: https://www.midi.org/specifications/item/xmf-specification-all

Non-File Formats

AES-3

AES3 (also-known-as AES-3 or AES/EBU) is not a file type. Some end users get confused about this when reading about this standard developed by the European Broadcast Union (EBU) and mistakenly believe there is such a thing. AES3 is a protocol for exchanging PCM audio data between audio hardware devices. It is not a file type.52

CD-Audio (.cda)

CD Audio track files (.cda) do not contain audio data. CDA files are designed only for use on physical medium (discs) and not intended for digital distribution such as via files or streaming. They are simply pointers - akin to a table of contents - advising an application where the audio data is found for a particular track on the current CD.

MIDI (.midi)

MIDI stands for Musical Instruument Digital Interface. It is a technical standard; not an audio codec or container. MIDI files are descriptive. That is, they describe characteristics specifically related to musical instruments and computers designed to play, edit, and record music.53 To be clear, MIDI files don't contain audio data. Think of them as metadata files only.

The original MIDI standard was released in 1983. In January 2019, MIDI version 2.0 was announced.54 Almost exactly a year later in January 2020, the new MIDI 2.0 was officially adopted.55

MIDI 2.0 Specifications

pdf

MIDI 2.0 Specification Overview

MIDI 2.0 specifications. Released 29 January 2020
Size: 298.98 KB
Date added: March 14, 2020
pdf

Common Rules for MIDI-CI Profiles

Part of MIDI 2.0 specifications. Released 29 January 2020.
Size: 314.80 KB
Date added: March 14, 2020
pdf

Common Rules for MIDI-CI PE

Part of MIDI 2.0 specifications. Released 29 January 2020.
Size: 954.77 KB
Date added: March 14, 2020
pdf

v1.1 MIDI-CI Specification

Part of MIDI 2.0 specifications. Released 29 January 2020.
Size: 1.04 MB
Date added: March 14, 2020
pdf

UMP and MIDI 2.0 Protocol Specification

Part of MIDI 2.0 specifications. Released 29 January 2020.
Size: 2.06 MB
Date added: March 14, 2020

Compression

This is an introductory discussion of audio file compression.

Audio compression refers to the compacting of audio data into a smaller file size, so that more information may be preserved in the same amount of data storage space.

All multi-channel codecs use audio compression to squeeze more than one channel of audio content into a single file.

Lossy Audio Compression

Lossy Audio Compression means a smaller file is created in exchange for the loss of some auditory data. It is a file compression technique whereby some audio data is discarded in order to increase the rate of file compression and reduce the overall file size. The discarded audio data is normally limited to sounds perceived to be outside the range of normal human hearing and/or within range but unlikely to be audible to most people.

Lossless Audio Compression

Lossless Audio Compression means higher audio fidelity in exchange for a larger file size. It is a file compression technique where a file compression algorithm is applied, but the original auditory data in the file is retained. Lossless audio compression results in larger file sizes compared with lossy audio compression methods.

Sometimes, a choice in lossless audio codecs is dependent on native operating system support. This chart provides a nominal cross reference:

macOS/iOS AIFC/AIFF (lossless), ALAC (lossy or lossless)
Linux AAC (lossy), MP3 (lossy), FLAC (lossless), Vorbis (lossy)
Windows AIFF (lossless), AAC (lossy), MP3 (lossy), WAVE (non-compressed)
Android MP3 (lossy), Ogg Vorbis (lossy), 3GP (lossy)

Which Codecs are Best?

When it comes to multimedia playing and recording, this is of course a loaded question. There is no clear winner under every circumstance. However, gleaning over numerous independent comparisons of various metrics does yield trends among some of the codecs described in this article. The reality is what is "best" is dependent on the task at hand.

Which Codecs Have the Best Audio Compression?

A high-level comparison of audio file compression efficiency. Smaller files indicate greater compression or greater recording length:file size efficiency. For example, a ratio of 100 KB to 1 second of 2-channel audio is more efficient than 200 KB of file size to 1 second of 2-channel audio. Therefore, a smaller file size means greater or more efficient audio file compression.

  • Average file size (smaller = better): Opus < Vorbis < MPC < AAC < MP3 < ALAC (lossy) < Monkey's Audio < FLAC < WavPack < ALAC (lossless) < AIFF < WAVe
  • Average file size Lossless Only: Monkey's Audio < FLAC < ALAC < AIFF < WAVe

Which Codecs Have the Best Audio Fidelity?

This varies quite a bit based on factors such as the recording bit rate. Furthermore, it is a very subjective measure. That said, the most important factor here is how codecs compare with one another, rather than which is "best." If you must have the best possible recording, you need to go with an uncompressed lossless format, no matter what other constraints you may have (i.e. a raw PCM or LPCM file such as Wave).

Bit rate is simply the level of granularity of detail that is preserved in the recording. Higher bit rates mean more information is retained. Higher bit rates use more storage capacity for the same length of time of recorded audio, as compared with lower bit rates. Therefore, given the same length of time that a song elapses, a higher bit rate file will be larger and take up more storage space, while the opposite is true of a lower bit rate recording.

Several independent studies appear to show the same or nearly the same relative rankings.76

  • Low (<128 kbps): Opus > ALAC > Ogg > WMA, AAC
  • Mid (128 kbps): Vorbis > AAC > MP3, MPC, WMA
  • High (>128 kbps): Vorbis, MPC > AAC, MP3

Which Codecs Are Faster?

Consumers are not normally concerned with the speed of audio codecs, as most of the time they are listening to music, and their music stream is likely buffered to begin with. However, under some circumstances the decoding speed in particular can be an issue. This is most apparent when streaming music over a mobile device, but this issue can become important under virtually any circumstance where network speed is slow or may potentially slow at any given moment.

  • Decoding speed (higher is faster/better):
    • WAVE > FLAC > MP3 > Vorbis > Opus > ALAC > Monkey's Audio (APE)

On the encoding/recording front, the relative speed of a codec often (but not always) is a good indicator of the amount of hardware resources required in order for it to run efficiently. Faster encoders tend to run more readily on limited hardware platforms, such as low-powered ARM processors for example. If your encoder is processing on a robust hardware platform, you will have greater freedom when choosing an encoder. However, bear in mind that more resource-intensive encoders will always take more time to encode the same original source material as compared with more efficient encoders. This is true regardless of the hardware platform of the encoding device.

  • Encoding speed (higher is better):
    • FLAC > WAVPack > Monkey's Audio (APE) > Vorbis > MP3 > Opus > AAC > ALAC

What is Metadata?

Metadata is any standardized technical or contextual information that describes the characteristics of content. For example, metadata could describe a song's title, author, and/or artist name. Metadata can be divided into these broad categories:

  1. Technical Metadata: a set of information that defines the essence structure, i.e. how the essence was edited and what source components were included in what derivation chain. Also known as "Structural Metadata."
  2. Contextual Metadata: a set of information that describes, parameterizes or catalogs content, such as episode number, copyright holder, etc. Also known as "Descriptive Metadata."
  3. Dark Metadata: is unknown to an application at the time of processing. This may be for many reasons including private metadata, unknown extensions to MXF and standardized metadata items that are not handled by the encoder.

Some codecs contain technical metadata, while generally speaking, codecs very rarely contain contextual metadata. Notable exceptions are MP3 and True Audio codecs. Most containers harbor technical metadata, and nearly all of them allow contextual metadata as well.

Bibliography

  1. AAC audio subtype. (6 April 2018). Internet Assigned Numbers Authority (IANA). https://www.iana.org/assignments/media-types/audio/aac
  2. AAC Decoder. (31 May 2018). Microsoft. https://docs.microsoft.com/en-us/windows/win32/medfound/aac-decoder
  3. "Advanced Systems Format (ASF) Specification; Revision 01.20.03." (December 2004). Microsoft Corporation.
  4. Angela. (18 December 2014). Files supported by Android. Online-Convert.com. https://blog.online-convert.com/files-supported-by-android/
  5. Android supported media formats. https://developer.android.com/guide/topics/media/media-formats
  6. Apple Core Audio Format Specification 1.0. (12 October 2011). https://developer.apple.com/library/archive/documentation/MusicAudio/Reference/CAFSpec/CAF_overview/CAF_overview.html
  7. Ashland, Matt. (12 March 2020). Monkey's Audio Version History. https://www.monkeysaudio.com/versionhistory.html
  8. Au file format. (4 September 2019). Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/Au_file_format
  9. Audio coding formats support. Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/Comparison_of_video_container_formats#Audio_coding_formats_support
  10. Brandenburg, K. (1999). MP3 and AAC Explained. Proceedings of the AES 17th International Conference: High-Quality Audio Coding, Florence, Italy, 1999 September 2-5.
  11. CAF File Overview. (12 October 2011). Apple Core Audio Format Specification 1.0. https://developer.apple.com/library/archive/documentation/MusicAudio/Reference/CAFSpec/CAF_overview/CAF_overview.html
  12. Codecs. (6 August 2008). Ubuntu Wiki. https://wiki.ubuntu.com/Media/Codecs
  13. Codecs list. Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/List_of_codecs
  14. File types supported by Windows Media Player. https://support.microsoft.com/en-us/help/316992/file-types-supported-by-windows-media-player
  15. FLAC format specification. Xiph.Org. https://xiph.org/flac/format.html
  16. Grill, B. and Quackenbush, S. (October 2005). MPEG-1 Audio. MPEG The Moving Picture Experts Group. https://mpeg.chiariglione.org/standards/mpeg-1/audio
  17. Harris, Mark. (14 January 2020). What Audio File Formats Can iPhone Play? Lifewire. https://www.lifewire.com/what-audio-file-types-can-iphone-play-2438433
  18. "Header file for Audio, .au" (10 November 1992). Sun Microsystems. https://pubs.opengroup.org/external/auformat.html
  19. Intro to Audio Programming, Part 2: Demystifying the WAV Format. (23 June 2009). Microsoft Corporation. https://docs.microsoft.com/en-us/archive/blogs/dawate/intro-to-audio-programming-part-2-demystifying-the-wav-format
  20. Intro to Audio Programming, Part 2: Demystifying the WAV Format. (22 June 2009). Microsoft. https://docs.microsoft.com/en-us/archive/blogs/dawate/intro-to-audio-programming-part-2-demystifying-the-wav-format
  21. Linear Pulse Code Modulated Audio (LPCM). (28 February 2017). Library of Congress. Sustainability of Digital Formats: Planning for Library of Congress Collections. https://www.loc.gov/preservation/digital/formats/fdd/fdd000011.shtml
  22. Matroska specifications. https://matroska.org/technical/specs/index.html
  23. Media container formats (file types). (12 December 2019). Mozilla Corporation. https://developer.mozilla.org/en-US/docs/Web/Media/Formats/Containers
  24. Monkey's Audio. (19 November 2019). hyrdogen audio. https://wiki.hydrogenaud.io/index.php?title=Monkey%27s_Audio
  25. Monkey's Audio. (22 December 2019). Wikipedia. https://en.wikipedia.org/wiki/Monkey%27s_Audio
  26. More Android audio format support info. https://developer.android.com/reference/android/media/AudioFormat
  27. MPEG-4 standard (metadata). https://en.wikipedia.org/wiki/MPEG-4_Part_14#Metadata
  28. Opus. http://opus-codec.org/
  29. Overview of the ASF Format. (31 May 2018). Microsoft. https://docs.microsoft.com/en-us/windows/win32/wmformat/overview-of-the-asf-format
  30. Pulse Code Modulation And Demodulation. (n.d.). ElProCus Technologies Pvt Ltd. https://www.elprocus.com/pulse-code-modulation-and-demodulation/
  31. Purnhagen, Heiko. (7 November 2001). MPEG Audio FAQ - MPEG-2: coded transmission/storage of sampled sound waves. https://sound.media.mit.edu/resources/mpeg4/audio/faq/mpeg2.html
  32. Quality vs Bitrate. https://opus-codec.org/comparison/
  33. Resource Interchange File Format (RIFF). 30 May 2018. Microsoft. https://docs.microsoft.com/en-us/windows/win32/xaudio2/resource-interchange-file-format--riff-
  34. Supported Audio File and Data Formats in OS X. (30 October 2017). https://developer.apple.com/library/archive/documentation/MusicAudio/Conceptual/CoreAudioOverview/SupportedAudioFormatsMacOSX/SupportedAudioFormatsMacOSX.html
  35. Supported media formats. (28 January 2020). Android Developer. https://developer.android.com/guide/topics/media/media-formats
  36. Tech Stuff - Audio/Video Files, Codecs and Containers. (21 December 2015). ZYTRAX. https://www.zytrax.com/tech/audio/formats.html
  37. True Audio (.tta) codec. http://tausoft.org/wiki/True_Audio_Codec_Overview
  38. van Rossum, Guido. (1 August 1994). Subject: FAQ: Audio File Formats (part 1 of 2). Internet Newsgroup. http://jedi.ks.uiuc.edu/~johns/links/music/audiofile1.html
  39. Vorbis Comment. (24 January 2020). Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/Vorbis_comment
  40. WavPack. (12 January 2020). Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/WavPack
  41. WebM Container Guidelines (video). https://www.webmproject.org/docs/container/
  42. What is LPCM (PCM). (n.d.). https://www.appgeeker.com/audio/what-is-lpcm.html
  43. What is the difference between M4A and AAC Audio Files? (7 August 2013). Stack Overflow. https://stackoverflow.com/questions/18110399/what-is-the-difference-between-m4a-and-aac-audio-files/18111039
  44. What’s True About Dolby® TrueHD? (2008) Dolby Laboratories. https://www.dolby.com/uploadedFiles/Assets/US/Doc/Professional/truehd-overview.pdf
  45. XMP SPECIFICATION PART 3: STORAGE IN FILES. 2016. Adobe Systems Incorporated. https://wwwimages2.adobe.com/content/dam/acom/en/devnet/xmp/pdfs/XMP%20SDK%20Release%20cc-2016-08/XMPSpecificationPart3.pdf
  46. Zibreg, Christian. (14 September 2017). iPhone X & 8 support FLAC playback, iOS 11 brings it to iPhone 7. iDownloadBlog.com (iDB). https://www.idownloadblog.com/2017/09/14/iphone-x-hardware-flac-codec/
  47. Zytrax. (21 October 2015). https://www.zytrax.com/tech/audio/formats.html

Endnotes

1 Linux requires PulseAudio for some Dolby and DTS codecs. PulseAudio is installed by default on most Linux distros.

2 Decoding capabilities only.

3 RealNetworks' RealAudio can use a separate file under some circumstances, with the file extension .ram, however these files appear to function as a combination of metadata and redirect to the actual audio file content. Therefore, in order to utilize metadata contained in the .ram file, it must be called first and the .ra file must not be called directly or there will be no metadata association.

4 Lossless by default with lossy "hybrid" mode. See WAVPack section in this document for more information.

5 Found only inside 3GP containers.

6 RIFF file type; follows RIFF metadata format.

7 Enhanced AC-3 files are supported natively by most modern operating systems, as an extension to Dolby Digital AC-3 support. However, while .eac3 files are known to exist (stand-alone audio files encoded in the E-AC-3 format), as a file extension/file type it is not natively supported by any operating system. If such files exist and their extensions are renamed to .ac3, they should work on an operating system that supports AC-3 natively.

8 Requires DirectShow Filters from Xiph.org to be pre-installed.

9 Limited support in iOS 11+ and macOS High Sierra (version 10.13; circa 2017). Opus codec ONLY compatible when inside CAF container.

10 Inside Matroska container only.

11 Audio codec. (31 October 2019). Wikipedia. https://en.wikipedia.org/wiki/Audio_codec

12 Joshi, Amrata. (22 October 2018). Opus 1.3, a popular FOSS audio codec with machine learning and VR support, is now generally available. Packt Publishing. https://hub.packtpub.com/opus-1-3-a-popular-foss-audio-codec-with-machine-learning-and-vr-support-is-now-generally-available/

13 Audio coding format. (22 November 2019). Wikipedia. https://en.wikipedia.org/wiki/Audio_coding_format

14 IMA = Interactive Multimedia Association. IMA is a now extinct organization that developed a set of audio algorithms, including a particular ADPCM algorithm.

15 Auro-3D. (20 December 2019). Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/Auro-3D

16 MPEG-4 Part 3 standard. Successor to MP3 (MPEG-2 Part 3 standard).

17 Plus up to 16 independent low-frequency range limited channels intended as sub-woofer channels.

18 Backwards compatible with the legacy AAC codec.

19 The number of channels supported by Adaptive Differential Pulse Code Modulation (ADPCM) codecs varies depending on the operating system and implementation, with a low limit of two (2).

20 Mac OS cannot encode MP3 files (see Supported Audio File and Data Formats in OS X (30 October 2017).

21 Only if inside a 3GP container.

22 Limited compatibility; decoding only.

23 Wouk, Kris. (8 August 2018). What is DTS:X? Digital Trends. https://www.digitaltrends.com/home-theater/what-is-dts-x/

24 After-market codec support is required. No operating system supports this codec natively.

25 Partial, inconsistent support for FLAC. Decoding supported in QuickTime. Playback not supported at all in iTunes. Spotty support in general with Mac OS and iOS.

26 iOS 11 or later only, but not in iTunes. Spotty support in general. Possibly impacted by hardware dependencies.

27 Decoding capability only.

28 Only inside a Mastroska container. Source: Supported media formats. (28 January 2020). Android Developers.

29 PCM = Pulse Code Modulation. In this context, the term is a broad definition that also refers to its variants such as Linear Pulse Code Modulation (LPCM).

30 The WAVE container supports up to 65,535 discrete PCM/LPCM/ADPCM channels.

31 Dolby Atmos was added to the TrueHD codec in 2016 (home theater systems). There is not an independent codec just for Atmos content.

32 Dolby Atmos does not actually incorporate channels. Instead, it supports up to 128 individual "objects." These sound objects require a device-specific decoder to be present or they will be ignored.

33 Requires a free add-on codec pack.

34 HE-AAC (High Efficiency Advanced Audio Coding) is also part of the original (MPEG-2) AAC standard; however, it is addressed independently in this article.

35 Wolters, Martin; Kjorling, Kristofer; Homm, Daniel; Purnhagen, Heiko. (13 October 2003). "A closer look into MPEG-4 High Efficiency AAC." Presented at the 115th Convention of the Audio Engineering Society, 10–13 October 2003.

36 iTunes Video and Audio Asset Guide 5.2. (2014). Cupertino, CA: Apple, Inc. Retrieved from https://www.apple.com/au/itunes/lp-and-extras/docs/iTunes_VideoandAudio_Asset_Guide5.2.pdf

37 Foresman, Chris. (October 28, 2011). After seven years, Apple open sources its Apple Lossless Audio Codec. Ars Technica. https://arstechnica.com/gadgets/2011/10/after-seven-years-apple-open-sources-its-apple-lossless-audio-codec/

38 Monkey's Audio. (8 February 2020). Retrieved from https://wiki.hydrogenaud.io/index.php?title=Monkey's_Audio

39 Dolby Digital Plus (DD+) is mentioned in RFC 4598, which was published in July 2006 by the IETF (Internet Engineering Task Force).

40 Search engine data analyzed from 2015-2020 using Google Trends (https://trends.google.com/trends/?geo=US

41 Don't allow yourself the become confused between Vorbis and Vorbis Comment. They are two different things. Their differences are explained in this article, under Vorbis Comment.

42 Cue sheets are formatted text files that provide index and other supplemental information for one or more audio files. They are called "cue sheets" partly becauase they are often used to create playlists (i.e. cueing a group of songs to play, in a particular order).

43 what is FLAC? (2014). Xiph.Org Foundation. https://xiph.org/flac/

44 Information sources: Musepack Source Code/Libs and Musepack.

45 Comparing the TTA codec with three popular lossless audio compressors. (9 November 2015). Tau Software. http://tausoft.org/wiki/True_Audio_Codec_Comparison

46 LFE = Low Frequency Emissions (i.e. bass frequencies; generally under 120 hz).

47 SLS is not technically a standard in and of itself; it is an extension of the AAC audio standard under MPEG-4 Part 3 (MPEG-4 Audio).

48 HD-AAC is a commercialized and proprietary variation of the Scalable Lossless Coding Standard (SDS) offered by the Fraunhofer Institute for Integrated Circuits. Functionally, HD-AAC is identical to SDS. Source: HD-AAC. (n.d.). https://www.idmt.fraunhofer.de/en/institute/projects-products/projects/hd-aac.html

49 MPEG-4 SLS. (2019). Via Licensing Corporation. https://www.via-corp.com/licensing/mpeg-4-sls/

50 MPEG-4 SLS. (2019). Via Licensing Corporation. https://www.via-corp.com/licensing/mpeg-4-sls/

51 μ-law (pronounced "u-law") is a telecommunications algorithm used to digitize and compress speech audio. It is applied predominantly in North America and Japan.

52 Limited support in Windows. FLAC in Matroska container is supported in Windows 10 only.

53 Experimental as of February 2020. See https://wiki.xiph.org/MatroskaOpus for more information.

54 Linux only.

55 Technically, WAVE (or WAV) is a Resource Interchange File Format (RIFF) container, and is modeled on the RIFF standard.

56 64-bit Resource Interchange File Format (RIFF) required for file sizes 4 GB or greater.

57 "Digital cellular telecommunications system (Phase 2+) (GSM); Universal Mobile Telecommunications System (UMTS); LTE; Transparent end-to-end packet switched streaming service (PSS); 3GPP file format (3GP) (3GPP TS 26.244 version 14.1.0 Release 14)." (January 2018). ETSI TS 126 244 V14.1.0 Release 14. Section 5.5: 16-17.

58 Any classification of AAC content qualifies (e.g. AAC-LE and HE-AAC, and HE-AACv2 to name a few).

59 The .aifc file extension denotes a compressed AIFF file.

60 Supports XMP (Extensible Metadata Platform), ID3v2 metadata (tags).

61 ASF (Advanced Systems Format). (6 April 2007). Library of Congress. https://www.loc.gov/preservation/digital/formats/fdd/fdd000067.shtml)

62 Au file format. (4 September 2019). Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/Au_file_format

63 G.711 is a PCM (Pulse Code Modulation) signal processing standard for voice frequencies. The G.711 PCM narrowband codec was created by ITU-T, which stands for the International Telecommunication Union Telecommunication Standardization Sector, an international body of experts that jointly develops interoperable telecommunications standards. The μ-Law portion of the standard relates to a particular variety of the standard which is found mostly in the United States and Japan. It is a very old standard (circa 1972), but is still used today. Source: "PULSE CODE MODULATION (PCM) OF VOICE FREQUENCIES: ITU-T Recommendation G.711." (1993, 1988, 1972). International Telecommunication Union.

64 Note the Part number change from the prior iteration, ISO/IEC 14496-12:2015 (Part 12).

65 The MPEG-4 multimedia standard does not have a defined set of structured metadata established by its standard. However, some MPEG-4 implementions do incorporate their own sets of pre-determined contextual metadata fields, and in these cases they are a form of pseudo-structured metadata. Regardless, the official standard does not have any.

66 "TECH 3306: RF64 - AN EXTENDED FILE FORMAT FOR AUDIO DATA." (June 2018). European Broadcasting Union. Version 2.

67 About the EBU. (2019). European Broadcasting Union (EBU). https://www.ebu.ch/about

68 Technical standards are backward compatible. However, playback implementation may not function properly if the player does not properly handle unfamiliar 64-bit WAVE/RIFF based chunks.

69 Support for ADPCM was not supported by WAVE's initial release in 1992, but was added later.

70 Reference: Shorten (file format). (4 January 2020). Wikipedia. Wikimedia Foundation. https://en.wikipedia.org/wiki/Shorten_(file_format)

71 Speex: A Free Codec For Free Speech. (2006). Xiph.Org. https://www.speex.org/

72 AES3 is a container standard applied only to the transmission of PCM audio data between professional audio devices. For an operating standard reference, see IEC 60958, Part 4.

73 Swift, Andrew. (May 1997), "A brief Introduction to MIDI", SURPRISE, Imperial College of Science Technology and Medicine.

74 Jeffrey, Cal. (29 January 2019). MMA and AMEI announce MIDI 2.0 prototyping has begun. TechSpot. https://www.techspot.com/news/78493-mma-amei-announce-midi-20-prototyping-has-begun.html

75 Details about MIDI 2.0™, MIDI-CI, Profiles and Property Exchange. (n.d.). The MIDI Association. https://www.midi.org/articles-old/details-about-midi-2-0-midi-ci-profiles-and-property-exchange

76 Lossless comparison. (28 September 2019). hydrogenaudio. https://wiki.hydrogenaud.io/index.php?title=Lossless_comparison#Comparison_Table