🎵 What does “compressing music” mean?
Compression means reducing the file size of an audio recording so it takes up less space (for example, turning a 50 MB song into a 5 MB MP3).
This is different from dynamic range compression, which adjusts loudness. Here, you’re asking about data compression (making files smaller).
🎯 How is music compressed? (Digital audio compression)
Most audio compression works by removing information that your ears don’t easily notice, using psychoacoustics—the science of how humans perceive sound.
🔑 Main Steps in Music Compression:
1️⃣ Splitting into frequency bands
-
The sound is broken down into lots of frequency slices (like a prism splitting light into colors).
-
This is called a transform, usually the Modified Discrete Cosine Transform (MDCT).
2️⃣ Psychoacoustic model finds what can be discarded
-
Your ear can’t hear quiet sounds that are masked by loud sounds nearby in frequency (called masking).
Example: a loud drum hit masks a faint hiss behind it.
-
The encoder analyzes which parts are likely inaudible and marks them to be removed or stored with lower precision.
3️⃣ Quantization
-
The remaining sounds are stored as numbers with fewer bits, reducing precision to save space.
-
Quiet or masked sounds get very coarse resolution.
4️⃣ Entropy coding
-
The encoder looks for repeating patterns and uses smart codes to store them efficiently (like zipping a text file).
For example, long strings of zeros are compressed into a short instruction: “repeat zero 100 times.”
5️⃣ Packing into frames
-
All this processed data is packaged into chunks called frames, with a bitstream header to tell the decoder how to reconstruct it.
📄 Example: MP3 Compression
An MP3 encoder uses all of the above steps.
-
Original CD-quality audio: 1.4 Megabits per second
-
Compressed MP3: 128–256 Kilobits per second, about 10x smaller
🎧 When you play it back:
-
The decoder reverses the steps: unpacks, decodes, reconstructs the frequency bands, and transforms back into time-domain audio.
-
Because some data was thrown out, this is called lossy compression.
💡 In Short: How music is compressed:
✅ Break sound into tiny frequency parts
✅ Remove or simplify things your ear won’t miss
✅ Store fewer bits for less important sounds
✅ Encode repeating patterns efficiently
✅ Reconstruct it later for playback
✨ Analogy:
Imagine describing a painting to a friend over the phone.
-
Instead of naming every brush stroke, you say:
“It’s mostly blue sky, with some white clouds here and there.” -
You’re conveying the essence but omitting tiny details.
That’s how MP3 and other codecs compress audio.