Audio encoding is the process of converting raw digital audio data into a specific file format, and understanding the fundamental differences between codecs is essential for making informed decisions about quality and file size. Raw uncompressed audio, as stored in WAV files, represents sound as a direct stream of PCM (Pulse Code Modulation) samples—numerical amplitude values captured at regular intervals. A stereo CD-quality WAV file at 44.1 kHz and 16-bit depth produces approximately 10 megabytes per minute of audio, which illustrates why compression codecs were developed.

Lossy compression codecs like MP3, AAC, and OGG Vorbis achieve dramatic file size reduction by exploiting psychoacoustic principles—the study of how humans actually perceive sound. These codecs use perceptual models to identify and discard audio information that most listeners cannot hear. For example, the phenomenon of auditory masking means that a loud sound at one frequency makes nearby quieter frequencies imperceptible, so the codec can safely remove those masked frequencies without noticeable quality loss. MP3 (MPEG-1 Audio Layer III), developed in the early 1990s, was the first widely adopted lossy codec and uses a modified discrete cosine transform to analyze and compress audio in frames of 1,152 samples. AAC (Advanced Audio Coding) is its successor, offering better quality at equivalent bitrates through improved spectral analysis and more efficient encoding of transient signals. OGG Vorbis is an open-source alternative that uses overlapping windowed transforms for smooth handling of varying audio content.

Lossless compression codecs like FLAC (Free Lossless Audio Codec) take a fundamentally different approach. Rather than discarding information, FLAC uses predictive modeling and entropy coding to reduce file size while preserving every single sample of the original audio. The encoder predicts each sample based on preceding samples, then stores only the small difference between prediction and reality, which requires fewer bits. This typically achieves 30-60 percent file size reduction compared to WAV while guaranteeing bit-perfect reconstruction of the original audio.

The concept of bitrate is central to lossy codecs and represents the number of bits allocated per second of audio. Higher bitrates allow the codec to preserve more spectral detail and reduce compression artifacts. At 128 kbps, MP3 encoding introduces audible artifacts on complex musical passages, particularly in high-frequency content like cymbals and vocal sibilance. At 256-320 kbps, these artifacts become virtually imperceptible to most listeners under normal conditions, a quality level often described as transparent. Understanding these trade-offs between format, bitrate, quality, and file size empowers you to choose the optimal encoding for your specific needs, whether archiving, streaming, or casual listening.