WAV vs MP3: Lossless vs Psychoacoustic Compression
WAV stores audio in its rawest digital form. A standard stereo CD-quality WAV file samples audio 44,100 times per second, stores each sample as a 16-bit integer, and does this for two channels simultaneously. The result: 44,100 × 16 × 2 = 1,411,200 bits per second, or roughly 10 MB per minute. A 4-minute song is a 40 MB WAV file.
MP3 takes a fundamentally different approach. Rather than storing every sample, it analyzes the audio using a psychoacoustic model — a mathematical representation of how human hearing works. The model identifies two categories of information: what you will certainly hear, and what you almost certainly won't notice. Sounds masked by louder simultaneous frequencies, very brief transients, and frequencies at the extremes of the audible range are discarded or severely quantized. Only the perceptually important information is stored.
The result of this process is a file 10–20 times smaller than the equivalent WAV, containing audio that sounds, to most human ears in most listening conditions, essentially identical. The "lossiness" is real — data is permanently discarded — but the loss is engineered to fall below the threshold of audibility.
Choosing the Right Bitrate
The bitrate of an MP3 is the number of bits per second used to represent the audio. Higher bitrates preserve more of the original audio data; lower bitrates sacrifice more to achieve smaller files.
At 128 kbps, the compression is aggressive enough that most listeners with decent headphones in a quiet environment can detect artifacts: a slightly "swirling" quality in reverb tails, reduced high-frequency definition on cymbals and sibilants, and a slight muddiness in complex musical passages. For background music in a video or a podcast, it's acceptable. For music you'll listen to critically, it's noticeable.
At 192 kbps, the encoding is transparent for the vast majority of listening scenarios. A double-blind ABX test comparing 192 kbps MP3 to the source WAV will stump most listeners most of the time. This is a reasonable default for everyday audio.
At 320 kbps, the encoding is considered reference quality. On consumer-grade headphones and speakers, 320 kbps MP3 is indistinguishable from the WAV source in virtually all blind tests. If you want the benefits of MP3 (smaller files, universal compatibility) without any perceptible quality penalty, 320 kbps is the ceiling to aim for.
Re-encoding Between Lossy Formats: Generational Degradation
Converting WAV to MP3 is a clean, first-generation lossy encode. You start with a lossless source and apply one round of psychoacoustic compression. Any artifacts introduced are minimal at appropriate bitrates.
The problem arises when you re-encode from one lossy format to another — or from MP3 back to MP3 at a different bitrate. Each encode discards different information and introduces its own artifacts. The artifacts from the first encode become part of the "source" for the second encode, which then adds its own layer of artifacts on top. After two or three re-encodings, the degradation can become clearly audible even at high bitrates.
A common mistake: receiving a 128 kbps MP3, wanting a higher-quality version, and re-encoding it to 320 kbps. The result is not a 320 kbps quality file — it's a 128 kbps quality file stored at a larger size, now with the artifacts from the original 128 kbps encode baked in permanently. Always encode from the highest-quality source available.
When to Keep the WAV
Music production is the clearest case for staying with WAV. Every time you apply an effect, mix tracks, or export from your DAW, you're performing mathematical operations on the audio data. These operations — particularly EQ, compression, and reverb — work best with the maximum precision that lossless audio provides. Converting to MP3 mid-project and then re-applying effects compounds the quality loss at each step.
Archiving original recordings is the other scenario. If you're preserving a live performance, a voice memo of historical interest, or a field recording for later use, the WAV is your master copy. You can always derive MP3 files from it later for distribution, but you cannot recover the discarded audio from an MP3. The WAV costs storage but preserves optionality.