Every time you stream audio over the internet, a codec is working invisibly in the background—compressing the raw audio signal from megabytes per second to kilobits per second, making real-time streaming over a mobile connection possible. The choice of codec affects not just audio quality, but compatibility with listeners' devices, server resource usage, and bandwidth requirements. This guide breaks down the four codecs supported by BroadcastMySelf and helps you make an informed decision for your station.
What Is an Audio Codec?
The word "codec" is a portmanteau of "coder-decoder." An audio codec is an algorithm that encodes raw audio data (PCM—Pulse Code Modulation) into a compressed format for storage or transmission, and decodes it back at the receiving end. Without codecs, streaming a 44.1 kHz stereo audio signal in its uncompressed form would require approximately 1.4 Mbps of continuous bandwidth—far too much for mobile broadcasting.
Codecs achieve compression through a process called perceptual coding. They analyze the audio signal and discard frequencies and details that human hearing is least sensitive to—frequencies masked by louder nearby sounds, frequencies above the typical hearing range, or subtle spatial cues that aren't perceptible in a monophonic playback context. The more aggressively a codec discards this data, the lower the bitrate—but also the more noticeable the compression artifacts become.
The goal of any broadcast codec choice is to find the minimum bitrate at which the compression artifacts are inaudible to your target audience on their likely playback devices. This is more nuanced than simply "higher bitrate = better."
MP3 (MPEG-1 Audio Layer III)
MP3 is the most universally recognized audio format in existence. Developed in the early 1990s and standardized by the Fraunhofer Institute, it became the dominant internet audio format through the late 1990s and 2000s and remains deeply embedded in the internet radio ecosystem today. The MP3 patent portfolio expired in 2017, making it fully open for use without licensing fees.
MP3 is the safest choice for internet radio compatibility. It is supported by every internet radio hardware player, every modern browser, every smart speaker, every car audio system with streaming capability, and essentially every software media player released in the last 25 years. If your goal is to reach the broadest possible audience without worrying about playback compatibility, MP3 is the default choice.
The quality characteristics of MP3 are well-understood: at 128 kbps stereo, it sounds excellent for most mixed content and is the historical standard for music distribution. At 64 kbps, quality is noticeably degraded for music but remains perfectly acceptable for speech. Below 48 kbps, the "underwater" artifacts of MP3's psychoacoustic model become clearly audible. For a music station, 192–320 kbps provides near-CD quality that is difficult to distinguish from the source on most playback systems.
AAC (Advanced Audio Coding)
AAC was designed as the technical successor to MP3, offering approximately 30% better compression efficiency at the same perceived quality level. A 128 kbps AAC stream generally sounds noticeably better than a 128 kbps MP3 stream, or equivalently, a 96 kbps AAC stream can match the perceptual quality of a 128 kbps MP3.
AAC is the native audio format of Apple devices and the iTunes ecosystem. It is natively supported on iOS, macOS, and in most modern Android devices and browsers. However, its support in older internet radio hardware players and legacy software is less universal than MP3. If your listeners are predominantly on modern smartphones or computers, AAC is a strong choice. If you have a significant audience on embedded internet radio hardware devices (dedicated tabletop internet radio players), check compatibility before switching.
For mobile broadcasting specifically, AAC-LC at 96–128 kbps offers an excellent quality-to-bandwidth ratio that is well-suited to the bitrate constraints of mobile data connections.
Ogg Vorbis
Ogg Vorbis is an open-source codec developed by the Xiph.Org Foundation—the same organization that created Icecast. It is designed as a free, patent-unencumbered alternative to MP3 and AAC. At equivalent bitrates, Ogg Vorbis generally offers better perceptual quality than MP3, and it is the native codec of the Icecast streaming ecosystem.
The primary limitation of Ogg Vorbis for internet radio is compatibility. While it is well-supported in desktop browsers (Firefox, Chrome) and VLC media player, it is not supported by Safari/iOS natively, and support in dedicated internet radio hardware players varies significantly. If you are certain your audience is using computer-based players or modern Android/Chrome devices, Ogg Vorbis at 96–128 kbps offers excellent quality. If you need broad compatibility including iOS and hardware devices, stick with MP3.
The Ogg container format (which wraps Vorbis audio data) can also be used to carry Opus audio, which brings us to the most modern option.
Opus
Opus is the most technically advanced audio codec available today and is the clear technical winner for internet radio in most modern deployments. Developed by the Xiph.Org Foundation and standardized by the IETF in 2012, Opus is specifically designed for real-time internet audio transmission—exactly the use case of live radio streaming. It covers the full quality range from 6 kbps (telephone quality) through 512 kbps (archival quality) with a single continuous codec architecture, unlike older codecs that require different profiles for different quality tiers.
Opus's advantage at low bitrates is dramatic. A 64 kbps Opus stream is perceptibly better than a 128 kbps MP3 stream for speech content. A 96 kbps Opus stream matches or exceeds 320 kbps MP3 for music content in many listening tests. This efficiency is particularly valuable for mobile broadcasters constrained by upload bandwidth.
The limitation, as with Ogg Vorbis, is compatibility. Opus is natively supported in all Chromium-based browsers, Firefox, and most modern Android and desktop media players. It is not supported natively in Safari on older iOS versions (support was added in iOS 16) or in most dedicated internet radio hardware. If your listeners are primarily on modern devices and you want the highest possible audio quality for a given bandwidth, Opus is the right choice.
Codec Comparison Table
| Codec | Best Bitrate Range | Compatibility | Strengths |
|---|---|---|---|
| MP3 | 64–320 kbps | Universal | Maximum compatibility, widely understood |
| AAC | 48–256 kbps | Excellent (modern devices) | 30% better efficiency than MP3 |
| Ogg Vorbis | 64–192 kbps | Good (desktop, Android) | Free, good quality, native Icecast format |
| Opus | 32–128 kbps | Good (modern browsers, Android) | Best efficiency, best for speech at low bitrates |
Which Codec Should You Choose?
The decision framework is straightforward. If you need to reach the widest possible audience with zero compatibility concerns, use MP3 at 128 kbps. This is the safe, universally understood choice that works everywhere. If your audience is primarily on modern devices and you want better quality for the same bandwidth cost, use AAC at 96 kbps. If you are using Icecast and your audience is on desktop browsers or Android, Ogg Vorbis at 96 kbps offers good quality with full Icecast integration. If you are optimizing for bandwidth efficiency and your audience is on modern devices, Opus at 48–64 kbps for speech or 96 kbps for music offers the best possible quality per bit.
BroadcastMySelf supports all four codecs. You can change your codec selection in Settings > Audio without needing to disconnect from your server. Note that changing the codec mid-broadcast will cause a brief stream interruption as the encoder reinitializes, so make codec decisions before going live.
A Note on Sample Rate and Channels
Codec choice is only part of the quality equation. Sample rate and channel configuration also matter. The standard sample rate for music is 44.1 kHz (matching CD quality). For speech-only content, 22.05 kHz or even 16 kHz is sufficient and reduces bandwidth by 50–64%. Mono encoding versus stereo also halves the required bitrate for the same quality level—for talk radio, mono at 64 kbps often sounds better than stereo at 64 kbps because the full bitrate is applied to a single channel rather than being divided between two.
Conclusion
Audio codec selection is one of the most consequential technical decisions a broadcaster makes, directly affecting both listener experience and bandwidth requirements. For most broadcasters in 2026, the practical choice is MP3 at 128 kbps for maximum compatibility, or Opus at 64–96 kbps if your audience skews toward modern devices and bandwidth efficiency is a priority. BroadcastMySelf makes switching between codecs simple, so experiment with your specific content type and measure the difference before committing to a long-term choice.