Lossless vs Lossy Compression
Lossless compression reduces file size without discarding any data — the original is perfectly reconstructed on decompression. Lossy compression permanently discards data to achieve smaller sizes, trading quality for smaller files.
Last updated
Lossless compression and lossy compression are the two fundamental approaches to reducing file size, each with completely different trade-offs.
Lossless Compression
The original data is perfectly reconstructed when the file is decompressed. No information is discarded.
Examples: ZIP, FLAC, PNG, TIFF (with LZW), ALAC, GZ
How it works: Identifies and eliminates redundancy in the data (e.g., repeated sequences, predictable patterns) without removing any actual content.
Best for: Data files (documents, archives, source code), audio mastering, image editing, and situations where quality cannot be compromised.
Typical ratios: 2:1 to 4:1 for most content
Lossy Compression
Data is permanently discarded to achieve compression. The original cannot be perfectly reconstructed.
Examples: MP3, AAC, JPEG, WebP, H.264, H.265, AVIF
How it works: Analyses what human senses are least likely to notice (psychoacoustics for audio, psychovisual models for images) and discards that data.
Best for: Distribution and sharing where file size matters more than perfect fidelity.
Typical ratios: 5:1 to 100:1 depending on quality settings
Generational Loss
The key danger with lossy compression: each re-encode multiplies the loss. Saving a JPEG at 70% quality, reopening it, and saving again at 70% applies compression twice — quality degrades each time. Always keep originals in lossless format and create compressed copies for distribution.