Lossless vs. Lossy Audio: WAV, FLAC, AAC, and MP3 Compared
Every audio file you stream, export, or record relies on an audio codec. Here is an architectural breakdown of psychoacoustic compression, mathematical differences, and when fidelity matters most.
The 3 Audio Format Categories
Uncompressed PCM
WAV, AIFF (1,411 - 4,608 kbps)
Raw binary representation of electrical voltage levels over time. Zero compression, zero mathematical processing overhead, highest disk space.
Lossless Compressed
FLAC, ALAC (600 - 900 kbps)
Like a ZIP archive for audio. Reduces file sizes by 40% to 50% using linear prediction and Huffman coding. Decodes to 100% bit-identical original PCM.
Lossy Perceptual
MP3, AAC, OGG (128 - 320 kbps)
Uses psychoacoustic algorithms to permanently discard sound humans cannot easily hear. Shrinks files by up to 90% for instant streaming and storage.
1. The Psychoacoustic Engine: How Lossy Encoders Save Space
Lossy compression codecs like MP3 (MPEG-1 Audio Layer III) and AAC (Advanced Audio Coding) do not simply chop off audio randomly. They run audio frames through a mathematical Modified Discrete Cosine Transform (MDCT) and evaluate them using biological models of human hearing:
Simultaneous Frequency Masking
When a loud sound and a quiet sound occur at frequencies close to each other at the exact same moment (e.g. a thunderous snare hit and a subtle room reverb harmonic), the human basilar membrane inside the inner ear is overwhelmed by the louder tone. The encoder removes the masked quieter sound completely.
Temporal Masking
For approximately 5 to 20 milliseconds before a loud transient attack (pre-masking) and up to 100 milliseconds after (post-masking), human hearing sensitivity drops sharply. Encoders allocate fewer bits or zero bits to sounds falling inside this window.
Absolute Threshold of Hearing (ATH)
Human ears are exquisitely sensitive between 1 kHz and 5 kHz (the range of human speech), but far less sensitive below 40 Hz and above 16 kHz. Encoders filter out extreme low sub-rumbles and gently roll off ultrasonic frequencies beyond 20.5 kHz.
2. Side-by-Side Format Comparison Matrix
| Format | Type | Typical Bitrate | 5-Min Song Size | DAW Editing Compatibility |
|---|---|---|---|---|
| WAV (16-bit / 44.1kHz) | Uncompressed | 1,411 kbps | ~50.4 MB | 100% Universal |
| WAV (24-bit / 48kHz) | Uncompressed | 2,304 kbps | ~82.4 MB | Studio Master Standard |
| FLAC | Lossless | ~850 kbps | ~30.0 MB | High (Modern DAWs) |
| MP3 (320 kbps) | Lossy | 320 kbps | ~11.5 MB | Preview only (avoids jitter) |
| AAC (256 kbps) | Lossy | 256 kbps | ~9.2 MB | Consumer playback (Apple Music) |
3. The Danger of Transcoding: Generational Loss
One of the most important rules in digital audio engineering is: never re-compress a lossy file into another lossy format.
If you take a 128kbps MP3 and export it as an AAC or another MP3, you are applying two consecutive psychoacoustic discarding passes. Each pass adds quantization noise, smears high-frequency transients, and creates watery artifacts.
Always preserve your original DAW project stems and master mixes in uncompressed 24-bit WAV format. Whenever you need an MP3 or AAC for sharing, create it directly from the pristine master WAV.
Frequently Asked Questions
Is AAC better quality than MP3 at the same bitrate?
Yes. AAC was developed as the successor to MP3. It utilizes larger filter bank frequency resolutions (up to 1,024 frequency lines compared to 576 in MP3), superior transient processing, and flexible joint-stereo coding. A 256kbps AAC generally matches or outperforms a 320kbps MP3 in fidelity.
Why don’t streaming services stream raw uncompressed WAV?
Raw WAV files consume approximately 10MB of cellular bandwidth every minute. For platforms serving tens of millions of concurrent streams, this represents massive server egress costs and frequent buffering on mobile networks. Codecs like FLAC or 256kbps AAC deliver indistinguishable listening fidelity at a fraction of the bandwidth.