Audio Compression Formats
Digital audio is stored and transmitted using a variety of formats, each balancing the trade-off between sound quality and file size. From the raw, uncompressed data used in professional studios to the highly efficient neural codecs used in modern communication apps, these formats define how we experience sound in the digital age.
Non-Compression Formats
Non-compressed audio preserves the original waveform without removing any data. The most common standard is Linear Pulse-Code Modulation (LPCM), often simply called PCM. This is the standard for Compact Disc Digital Audio (CD-DA) and is typically stored in container formats like WAV, AIFF, or AU.
Other specialized non-compression methods include Pulse-Density Modulation (PDM), which is the basis for Direct Stream Digital (DSD) used in Super Audio CDs, and Pulse-Amplitude Modulation (PAM).
ไม่มีภาพประกอบ
Lossless Compression
Lossless compression reduces file size without losing any original audio data. When decoded, the audio is bit-for-bit identical to the original source.
Popular Lossless Codecs
- FLAC (Free Lossless Audio Codec): One of the most widely used open-source lossless formats.
- ALAC (Apple Lossless Audio Codec): Apple's proprietary lossless format used in Apple Music and QuickTime.
- WavPack (WV) and Monkey's Audio (APE): Other common alternatives for high-fidelity archiving.
Specialized and Professional Lossless Formats
In the cinema and DVD industry, formats like Dolby TrueHD (based on MLP) and DTS-HD Master Audio are used for high-definition surround sound. Other technical standards include MPEG-4 ALS (Audio Lossless Coding) and SLS (Scalable Lossless Coding).
Lossy Compression
Lossy compression achieves significantly smaller file sizes by permanently removing audio information that the human ear is less likely to perceive. Most of these codecs rely on mathematical transforms like the Discrete Cosine Transform (DCT) or the Modified Discrete Cosine Transform (MDCT).
General Purpose Audio
- MP3 (MPEG-1 Audio Layer III): The most famous lossy format, though now joined by more efficient successors.
- AAC (Advanced Audio Coding): The standard for YouTube, iTunes, and MPEG-4, offering better quality than MP3 at similar bitrates.
- Vorbis and WMA (Windows Media Audio): Alternative lossy formats used across various platforms.
Neural Audio Codecs
The latest evolution in audio compression involves artificial intelligence. Neural audio codecs use machine learning to compress audio far more aggressively than traditional methods. Examples include Google's Lyra (used in Google Duo), Microsoft Teams' Satin, and Facebook's EnCodec.
Specialized Audio Applications
Bluetooth Audio Codecs
Bluetooth devices use specific codecs to transmit audio wirelessly. While SBC (Low Complexity Subband Coding) is the universal standard, high-resolution options include Sony's LDAC, Qualcomm's aptX HD, and the newer LC3 used in Android 13.
Speech and Voice Codecs
Voice-optimized codecs focus on the human vocal range to maximize clarity at very low bitrates. These often use Linear Predictive Coding (LPC). Common standards include the ITU G-series (like G.711 for telephony) and Opus, a versatile IETF standard used for VoIP.
Gaming and Mobile Hardware
Video game consoles often use proprietary ADPCM (Adaptive Differential Pulse-Code Modulation) variants. For example, Nintendo uses GCADPCM, while Sony uses ATRAC9 for the PS4 and PS Vita.
Key Facts
- LPCM is the gold standard for uncompressed audio (CDs).
- FLAC and ALAC are the primary choices for lossless compression.
- AAC generally provides superior efficiency and quality compared to MP3.
- Opus is a leading modern standard for low-latency speech and general audio.
- Neural codecs (e.g., Lyra, EnCodec) represent the cutting edge of AI-driven compression.
| Type | Data Retention | File Size | Common Examples |
|---|---|---|---|
| Uncompressed | 100% (Exact) | Very Large | PCM, WAV, DSD |
| Lossless | 100% (Reconstructed) | Medium | FLAC, ALAC, WavPack |
| Lossy | Partial (Perceptual) | Small | MP3, AAC, Vorbis |
| Neural | AI-Synthesized | Very Small | Lyra, Satin, EnCodec |
Frequently Asked Questions
What is the difference between lossless and lossy compression?
Lossless compression reduces file size without losing any data; the original audio can be perfectly reconstructed. Lossy compression removes "unnecessary" data to achieve much smaller files, meaning the original audio cannot be perfectly restored.
Which is better: MP3 or AAC?
AAC is generally considered superior to MP3 because it uses more advanced compression techniques, resulting in better sound quality at the same bitrate.
What are neural audio codecs?
Neural audio codecs use neural networks and machine learning to compress audio. They can maintain intelligible speech and sound at extremely low bitrates that would be impossible for traditional codecs.
What is PCM audio?
Pulse-Code Modulation (PCM) is a method used to digitally represent analog signals. It is the most basic form of uncompressed digital audio used in WAV files and CDs.
Why are there so many Bluetooth codecs?
Different manufacturers develop codecs to improve specific aspects of wireless audio, such as reducing latency (aptX Low Latency) or increasing fidelity for high-resolution audio (LDAC, aptX HD).