Overview of Audio File Formats

Last updated on April 23, 2026 by Sören Ramspeck with the experience of more than 128,379,858 converted files since 2013.

Anyone looking to convert, play, or archive audio files is faced with an astonishing variety of formats. Whether it’s MP3, AAC, WAV, or OPUS, the list is long and the differences are sometimes very subtle. Our guide explains all relevant audio file formats—what they were designed for, where they’re used today, and which format is the best choice for which use case.

Basics: Codecs, Containers, and Compression

Readers who are already familiar with these terms are welcome to skip this section and proceed directly to the music formats.

Codec

A codec (coder/decoder) refers to the process used to compress and decompress audio data (and also video data). MP3, Vorbis, and FLAC are codecs.

Container

A container, on the other hand, is the file format that “wraps” or “holds together” one or more codecs. For example, an M4A file is an MPEG-4 container that can contain AAC audio or Apple Lossless. The file extension therefore does not always directly reveal the codec used, but rather indicates the container used.

Compression

There are three main approaches to compression:

Compression type	Description	Examples
Lossy	Removes audio components that the human ear can barely perceive or not at all. This results in smaller files, but the removed information is irretrievably lost.	MP3, AAC, OGG Vorbis, Opus
Lossless	Compresses the data without losing any information, like a ZIP file for audio. The original data can be restored exactly.	FLAC, ALAC, APE, WavPack
Uncompressed	Raw PCM audio data without any compression. Maximum quality, but also maximum file size.	WAV, AIFF

Music formats (lossy)

Lossy music formats are what most people encounter on a daily basis: when listening to music, streaming, or downloading.

MP3 – arguably the most well-known music format

MP3 (MPEG-1 Audio Layer III) is the world’s most popular audio format. Developed in the late 1980s and early 1990s at the Fraunhofer Institute for Integrated Circuits (IIS) in Erlangen, MP3 was released in 1993 as part of the MPEG-1 standard.^[1] MP3 achieved its real breakthrough in the late 1990s with the first portable MP3 players. Since then, MP3 has practically become synonymous with digital music.

In April 2017, the licensing program for MP3 run by Technicolor and Fraunhofer IIS was discontinued, and the last core patents have expired.^[2] The format has since been completely royalty-free to use, which is another reason for its extremely widespread use to this day.

Typical bitrates range from 128 to 320 kbps. Modern encoders such as LAME (libmp3lame) deliver a quality at 192 kbps and above that, in blind tests, most listeners can no longer distinguish from CD quality. ^[3]

The days when MP3s sounded noticeably "tinny" are over with current encoders, even though prejudices against MP3 persist among some audiophiles and advocates of lossless formats.

AAC – A Powerful Contender to MP3

AAC (Advanced Audio Coding) was adopted in 1997 by a consortium including Fraunhofer IIS, Dolby, Sony, and AT&T as part of the MPEG-2 standard. AAC finally made its breakthrough in 2003 when Apple ^[4]launched the iTunes Store and chose AAC as its standard music format. Today, AAC is the standard in MP4 containers, on Apple devices, on YouTube, and on many streaming platforms.

AAC is often referred to as the "successor" or "replacement" for MP3. Technically speaking, AAC does use a more modern psychoacoustic model and more efficient compression methods^[5]; in practice, however, the difference in quality is smaller than is often portrayed. At bitrates relevant in practice—starting at 128 kbps—AAC and MP3 are virtually indistinguishable when using modern encoders. Many older comparisons are based on MP3 encoders that perform significantly worse than today’s LAME. The claim from that era likely persists to this day. At very low bitrates below 96 kbps, however, AAC still offers measurable advantages thanks to techniques like SBR (Spectral Band Replication). Yet such low bitrates are rarely needed in practice anymore.

Unlike MP3, AAC is still subject to patent protection. The last basic patents are expected to expire in 2028, while some extensions, such as HE-AAC^[6], will not expire until 2031. AAC files are often found in M4A containers or as RAW AAC files.

OGG Vorbis – the open alternative

OGG Vorbis is a completely open and patent-free audio format developed by the Xiph.Org Foundation. "Vorbis" refers to the codec, and "OGG" refers to the container. The first stable version (1.0) was released in July 2002^[7]. Development began as a direct response to the MP3 licensing fees at the time.

Technically, Vorbis delivers solid quality at comparable bitrates. However, the format never gained traction in the consumer market; it lacked hardware support or a major corporation to back it on the hardware side. Nevertheless, Vorbis has found a permanent home at Spotify, as Spotify uses OGG Vorbis at bitrates of 96, 160, and 320 kbps on both desktop and mobile apps.^[8] Vorbis is also popular in modern video games because it can be integrated without a license.

Opus – the latest lossy audio codec

Opus is the most technically versatile and modern lossy audio codec. Opus was standardized^[9] by the IETF in 2012 as RFC 6716 and is a joint project of Xiph.Org (like Vorbis) and Microsoft (formerly Skype). The codec is also completely open-source and royalty-free.

What makes Opus special is that the codec can handle an extremely wide range, from narrowband speech at 6 kbps to high-resolution music at 510 kbps. It seamlessly switches between a speech encoder (based on SILK, developed by Skype) and a music encoder based on CELT.^[9] This versatility makes Opus the standard in WebRTC, the core technology behind browser-based telephony and video conferencing.^[10] WhatsApp, Signal, Discord, and Telegram also use Opus for voice messages and calls.

Despite its technological superiority, Opus has not become the new MP3. It lacks visibility in the consumer sector, and most users encounter Opus without even realizing it: when using the aforementioned apps or watching WEBM videos on platforms like YouTube.

WMA – Microsoft’s Approach with DRM

WMA was Microsoft’s attempt to establish its own lossy audio format. WMA was introduced in 1999 and delivered quality comparable to MP3 at the same bitrates. What particularly pleased the music industry, however, was that WMA offers integrated DRM (Digital Rights Management). This was intended to prevent unauthorized sharing. Microsoft’s Zune players relied on WMA but were unable to truly establish themselves in the market.

Today, WMA has become practically irrelevant. Even Microsoft has abandoned the format in its current products in favor of AAC and MP3. Today, WMA files are mostly found in older music collections or voice recordings.

MP2 – MP3’s Parent

MP2 (MPEG-1 Audio Layer II) is the direct predecessor of MP3 and was also developed by Fraunhofer IIS. Although MP2 has long since been replaced by MP3 in the consumer sector, the format remains a standard in broadcasting, as DAB radio and DVB television audio still use MP2 to this^[11] day. Why? Because MP2 is more resilient to interference and transmission errors, which, as is well known, cannot be avoided in radio broadcasting.

AC3 (Dolby Digital) – Cinema Standard

Strictly speaking, AC3 is not actually a pure audio format, but a multichannel audio codec from Dolby Labs. It supports up to 5.1-channel surround sound and is the standard audio codec on DVDs, many Blu-ray movies, and in digital television (replacing MP2).^[12] Numerous streaming services also use Dolby Digital or its successor EAC3 (Enhanced AC3, also known as Dolby Digital Plus) for their multichannel audio tracks. Pure .ac3 files are very rare; the codec is mostly used in video containers such as MKV or MP4.

Comparison of Lossy Music Formats

Audio Formats

Speech and music place fundamentally different demands on the audio codec used. While human speech uses only the frequency range of about 200 to 4,000 Hz, music typically requires 20 to 20,000 Hz. Speech codecs can specifically exploit this narrower range to deliver intelligible results even at very low bit rates. Added to this is often the requirement for low delay (latency), such as in telephony or video conferencing. Those who are a bit older will surely still remember poor telephone connections where people kept interrupting each other.

AMR – GSM Mobile Communications Standard

AMR (Adaptive Multirate) was developed by ETSI for GSM mobile networks and adopted as a 3GPP standard in 1999.^[13] With a frequency range of 200 to 3,400 Hz,AMR is narrowband and, as the name implies, can dynamically adapt to the available bandwidth between 4.75 and 12.2 kbps. This is perfectly sufficient for voice, but completely unsuitable for music.

Android devices and older iOS versions use AMR as the standard for voice recordings. MMS voice messages and some VoIP apps also use AMR. However, AMR files should be converted to MP3 or AAC for sharing, as not every player supports AMR.

3GP and 3GPP – Widespread Use

3GPP (Third Generation Partnership Project) is a container format developed specifically for 3G mobile phones and released in 2003. 3GPP is optimized for limited storage space and slow connections.^[14] Technically, it is based on the MPEG-4 Part 12 container and can use the AMR or AAC codecs for audio.

Although 3GPP is primarily a video format, it is often encountered as a pure audio file. In Android’s MediaRecorder API, the `OutputFormat.THREE_GPP` with the `AMR_NB` encoder is documented^[15] as the standard example for audio recordings, and many older and more affordable Android devices still use this format today for voice recordings. Samsung also long used the .3ga (3rd Generation Audio) file extension for its Voice Recorder app.^[16] In the MMS realm as well, .3gp files remain widely used as containers for voice and short video clips.

However, most 3GPP files in circulation today originate from older mobile phones. For sharing, it is recommended to convert them to MP3 for audio-only recordings or MP4 for videos. VLC and Windows Media Player can generally play .3gp files without any issues.

QTA (QuickTime Audio) – Apple’s New Development

QTA is Apple’s latest audio format, introduced in late 2024 with iOS 18.2 for the iPhone 16 Pro and Pro Max. QTA serves as an alternative to the previous .m4a format for voice memos on these devices and introduces a fundamental innovation: spatial audio.

Technically, QTA is based on First Order Ambisonics (FOA). The iPhone’s four microphones (starting with the iPhone 16) capture a three-dimensional sound field that is stored in four “channels.” The codec used for this is called APAC (Apple Positional Audio Codec). APAC is a new, proprietary codec that Apple originally developed for immersive video on the Vision Pro (Apple’s VR headset).^[17] QTA files also include a stereo AAC track as a fallback for older devices.^[18] The spatial audio data allows background noise to be better isolated and removed later, for example in Logic Pro.^[19]

QTA shows where the future is headed: away from simple mono or stereo recordings, toward spatial audio that can be edited even more effectively afterward and adapted to the listener’s position.

Opus as a speech codec

I’ve already introduced Opus in the music formats section, but it deserves a separate mention in the voice category. This is where Opus really shines: the codec delivers excellent voice quality at 12–20 kbps while also having low enough latency for real-time communication. Opus is now the standard voice codec in Web^[10]RTC and thus in all browser-based telephony and video solutions, including WhatsApp, Signal, Discord, Telegram, and many other applications. It is no exaggeration to say that Opus is the most widely used voice codec on the internet, even though very few users have likely ever heard of it.

Lossless and uncompressed formats

Lossless formats store audio without any loss of quality. This means the original data can be reconstructed with bit-for-bit accuracy. Unlike MP3 or AAC, nothing is lost. The trade-off, however, is significantly larger file sizes. Uncompressed formats like WAV generate the largest amounts of data.

FLAC – free and lossless

FLAC (Free Lossless Audio Codec) is by far the most popular lossless audio codec. FLAC (like Opus and Vorbis) was developed by the Xiph.Org Foundation and has been available since 2001; it is completely open-source and royalty-free.^[20] Depending on the content, FLAC compresses audio losslessly to about 50–70% of the original size.

FLAC has established itself as the standard in the audiophile scene and is used by streaming services such as Tidal, Qobuz, and Amazon Music for lossless streaming. Since September 2025, Spotify has also offered lossless streaming in the FLAC format (24-bit / 44.1 kHz for Premium users).^[21] Android supports FLAC natively. Only Apple has resisted FLAC and uses ALAC:

ALAC – free and lossless with an “Apple flavor”

ALAC (Apple Lossless Audio Codec) is Apple’s answer to FLAC. Introduced in 2004, ALAC delivers comparable compression and identical quality to FLAC; after all, both are lossless. Since 2011, ALAC has been open source^[22], which ultimately reduces the technical differences from FLAC to the container format. ALAC is packaged in Apple’s .m4a container and is seamlessly integrated into iTunes, Apple Music, and other Apple devices. If you want to listen to lossless audio within the Apple ecosystem, there’s practically no way around ALAC. Apple Music uses ALAC for its “Lossless” offering.^[23] So ALAC is basically FLAC with an Apple twist 😉 and plays no role outside the Apple world, where FLAC is the first choice.

WAV – A Veteran of Audio Files

WAV (Waveform Audio File Format) was introduced in 1991 by Microsoft and IBM and stores uncompressed PCM (Pulse Code Modulation) audio data. Due to the simplicity of the format, it places no great demands on processing power and was the standard for the built-in audio recorder in earlier versions of Windows. Even today, WAV remains the standard exchange format in recording studios and professional audio production. The quality is top-notch, as there is absolutely no compression. However, the files are correspondingly large: around 10 MB per minute at CD quality (44.1 kHz, 16-bit, stereo).

Due to its file size, WAV is impractical for archiving or transferring music. FLAC is the better choice here: same quality, but only about half the file size.

AIFF

AIFF (Audio Interchange File Format) is Apple’s equivalent to WAV. Developed by Apple in 1988, AIFF is also based on uncompressed PCM data. In the studio environment on the Mac, AIFF was long the standard choice, but is now increasingly being replaced by FLAC or ALAC. Technically, there is no difference in sound quality between WAV and AIFF, as both store the same raw data, just in different container structures.

APE and WavPack

In addition to FLAC and ALAC, there are other lossless codecs that play a role in certain niche markets. APE (Monkey’s Audio) achieves a slightly better compression rate than FLAC, but is significantly more computationally intensive and is supported by only a few players. APE is occasionally encountered in audiophile forums, particularly in Asia.

WavPack offers an interesting feature: a hybrid mode that stores a lossy and a lossless layer in separate files. If you keep only the .wv file, you have a compact lossy version. If you add the .wvc correction file, you can reconstruct the original with bit-perfect accuracy.^[24] In practice, however, WavPack is just as rare as APE.

Lossless Formats Compared

Containers and file extensions: the big misunderstanding

One of the most common misconceptions about audio formats: the file extension does not necessarily reveal which codec is inside the file. The reason for this is the container principle, in which the container is merely the outer shell that contains the codec.

M4A – one container, two worlds

M4A is an MPEG-4 audio container and the best example of this confusion: An .m4a file can contain lossy AAC audio or lossless ALAC audio. This cannot be determined from the outside. Only a look at the metadata (e.g., using ffprobe or MediaInfo) reveals the actual codec. Apple consistently uses .m4a for both variants, which regularly leads to confusion when users believe their entire iTunes library is “lossless,” even though most files contain AAC, or files suddenly become unplayable on another device.

The OGG Family

The history of OGG file extensions is an instructive example of how well-intentioned standards can clash in practice. Originally, .ogg was used for all content in the Ogg container. But when Theora video and other codecs were added to the container alongside Vorbis audio, the Xiph.Org Foundation introduced differentiated extensions in 2008 with RFC 5334:^[25]

In practice, this diversity continues to cause problems to this day. A current example: Android 15 generates .OGX files when downloading .ogg files if the web server sends the outdated MIME type application/ogg instead of audio/ogg.^[26] The files themselves are perfectly fine; only the extension is incorrect. Among others, Wikimedia and Fandom/Wikia were affected. Details on this in the analysis: The sudden appearance of OGX files.

WebM

WebM is Google’s open multimedia container based on the Matroska format (MKV).^[27] For audio, WebM uses either Vorbis or Opus as a codec, and YouTube uses WebM/Opus for its audio streams. Pure WebM audio files are rarely encountered in everyday use; the container is primarily encountered in a web context.

Special Audio Formats

MIDI

MIDI (.MID) is an exception: A MIDI file does not contain actual audio data, but rather control instructions. Essentially, "notes for the computer." Which notes are played, how loud, and for how long is stored, and how it ultimately sounds depends on the playback device and its sound generation. While an MP3 file always sounds the same, MIDI is a matter of interpretation, just like a sheet of music that is played differently by various musicians.

Today, MIDI is primarily used in music production, where it serves as a universal interface between software instruments, synthesizers, and DAWs (Digital Audio Workstations). The current MIDI 2.0 specification was adopted in 2020 and expands the standard to include higher resolution and bidirectional communication.^[28] .mid files are also common in the karaoke sector: in addition to the sheet music, they often include the song lyrics.

DSD (DSF / DFF)

DSD (Direct Stream Digital) is a completely different approach to digital audio recording. While all formats mentioned so far are based on PCM sampling (i.e., amplitude values at regular intervals), DSD uses 1-bit sigma-delta modulation at extremely high sampling rates (2.8 MHz and higher; for comparison, CD audio operates at 44.1 kHz).

DSD was developed by Sony and Philips for the Super Audio CD (SACD).^[29] Whether DSD actually sounds audibly better than high-resolution PCM is a matter of debate among audiophiles. One thing is certain: the files are very large and software support is limited. DSD files have the extension .dsf or .dff and are found exclusively in the audiophile niche.

RealAudio

RealAudio (.ra, .ram) was the dominant format for audio streaming on the Internet in the late 1990s and early 2000s, and RealPlayer was installed on countless computers. Today, RealPlayer has long been discontinued (thank goodness!), RealAudio is completely obsolete, and has been replaced by more efficient codecs and more universal streaming protocols. You only encounter .ra files in old files anymore.

CAF (Core Audio Format)

CAF is Apple’s universal audio container, introduced with macOS 10.4 (2005). CAF can accommodate virtually any audio codec and has no file size limit, unlike WAV, which hits the limit of the RIFF header at 4 GB.^[30] In practice, .caf files are encountered almost exclusively in Apple development and iOS programming.

Summary

Complete overview of all audio formats

Why do we use what we use?

To understand the diversity of formats, one must understand the ecosystems behind them. The prevalence of an audio format depends only to a limited extent on its technical quality; timing, patents, and the market power of the companies supporting a format are far more decisive.

Why is MP3 still dominant?

MP3 had a head start of nearly a decade over its alternatives. By the time AAC and OGG Vorbis appeared, MP3 was already the universal standard, and every device, every piece of software, and every operating system could play MP3. On top of that, MP3 has been completely patent-free since 2017.^[2] There is simply no compelling reason to switch to a different format.

Apple: The Closed Ecosystem

Apple consistently relies on its own or Apple-preferred formats: AAC for lossy music (iTunes Store, Apple Music), ALAC for lossless streaming (Apple Music Lossless)^[23], and, since late 2024, QTA with the proprietary APAC codec for spatial audio recordings.^[17] There’s a method to this strategy: Apple controls the entire stack, from the codec to the container to the end device. This guarantees seamless integration and binds users to the ecosystem. It is telling that Apple still does not natively support FLAC, even though there is technically no reason not to.

Google: the open approach

Google pursues a different strategy and relies on open formats such as OGG Vorbis and Opus in the WebM container.^[27] YouTube uses Opus for its audio streams, and Android natively supports FLAC, Opus, and Vorbis. Google even actively funds the development of Opus and drives its adoption via WebRTC.^[10] The downside of the open approach: it lacks the tight vertical integration that Apple offers. Not every Android device handles all open formats equally well, and the fragmentation of the Android ecosystem regularly causes compatibility issues, as the OGX example on Android 15 demonstrated.^[26]

Streaming Services: Who Uses What?

The streaming table above clearly illustrates the pattern: Apple relies on AAC and ALAC, Google on Opus, and the rest of the industry fluctuates between AAC and FLAC. Spotify was long a special case with OGG Vorbis as its only lossy format, but in 2025 it followed suit with FLAC for lossless playback.

Conclusion: Which format for which purpose?

There is no "best" audio format, but there is a sensible choice for every use case:

The world of audio formats has grown over the decades and will continue to do so. New developments like Apple’s QTA show that the journey is far from over, as spatial audio and adaptive codecs are sure to gain even more importance in the coming years. For everyday use, however, MP3 and FLAC are your safe and compatible choices.

Sources

[1] Fraunhofer IIS: mp3 – Overview and History
[2] Fraunhofer IIS Audio Blog: Alive and Kicking – mp3 Software, Patents and Licenses (2017)
[3] LAME MP3 Encoder
[4] Fraunhofer IIS: MP3 and AAC Explained (PDF, AES-17 Conference)
[5] Wikipedia: Advanced Audio Coding – Technical Details
[6] Via Licensing (formerly Access Advance): AAC Patent Portfolio
[7] Xiph.Org Foundation: Vorbis – Open Audio Codec
[8] Spotify Support: Audio Quality
[9] RFC 6716: Definition of the Opus Audio Codec (IETF, 2012)
[10] W3C: WebRTC 1.0 – Real-Time Communication Between Browsers
[11] ETSI: Digital Audio Broadcasting (DAB) Standards
[12] Dolby: Dolby Digital – Technology Overview
[13] 3GPP: Media Codecs – AMR Speech Codec
[14] 3GPP: Specifications – 26 Series (Codecs and Media Formats)
[15] Android Developers: MediaRecorder.OutputFormat – THREE_GPP
[16] Samsung Community: Voice Recorder and 3gp4 Codec (M4A/3GA)
[17] Apple Developer: Apple Positional Audio Codec (PDF)
[18] file-converter-online.com: QTA File Format – ffprobe Analysis
[19] Apple Support: Apply Spatial Playback Processing to QTA Files in Logic Pro
[20] Xiph.Org Foundation: FLAC – Free Lossless Audio Codec
[21] Spotify for Artists: Audio File Formats (FLAC Support)
[22] Apple Lossless Audio Codec – Open Source (GitHub/macOS Forge)
[23] Apple Support: About Lossless Audio in Apple Music
[24] WavPack: Technical Documentation – Hybrid Mode
[25] RFC 5334: Ogg Media Types (IETF, 2008)
[26] file-converter-online.com: The Sudden Emergence of OGX Files – An Analysis (2026)
[27] The WebM Project – An Open Web Media Project
[28] MIDI Association: Details About MIDI 2.0
[29] Sony: Super Audio CD – Overview
[30] Apple Developer Documentation: Core Audio Format