What lossy compression removes
Lossy codecs represent audio with fewer bits by discarding or approximating information according to a perceptual model. What is removed depends on codec, encoder, settings, and source. The process can be effective for delivery, but it is not reversible: decoding produces playable samples, not the exact original master.
A bitrate estimate uses duration × average bitrate. It predicts payload size, not audible quality. The page never assigns a “quality percentage,” because a number like 80% would hide encoder and content differences.
What lossless compression preserves
A lossless codec such as FLAC can reconstruct the same PCM sample values presented to its encoder. Its output size depends on how predictable the audio is, so a fixed ratio is not an exact law. The calculator shows an adjustable conservative range instead of claiming one result for silence, speech, dense music, and noise.
Lossless does not mean that an earlier recording or encode was perfect. If MP3 is decoded and saved as FLAC, FLAC preserves that decoded signal but cannot recover information removed before it received the samples.
WAV, FLAC, MP3, AAC and OGG
WAV commonly contains uncompressed PCM, so its audio data is determined by duration, sample rate, bit depth, and channels. FLAC compresses PCM without changing decoded sample values, but its ratio is content-dependent. MP3, AAC, and OGG/Vorbis are lossy delivery families whose size is commonly planned from average bitrate.
Containers and codecs are related but not identical concepts. This site’s browser tools expose only the audited combinations in their capability matrix. Read Sample Rate vs Bitrate for the two underlying size models.
Editing, archiving and sharing workflows
Keep a lossless master from the original capture or highest-quality production source when future editing matters. Repeated lossy editing and export can accumulate changes. For sharing, streaming, a website, or a feed, create a smaller lossy delivery copy from that master and keep the roles separate.
An archive decision also includes metadata, file integrity, backups, documentation, and format support—not merely compression ratio. A delivery copy is replaceable; a unique original may not be.
Why lossy to lossless does not restore detail
Transcoding MP3 to FLAC decodes the MP3 and stores those decoded samples losslessly. It may prevent an additional lossy encoding step in a later process, but it cannot infer the exact samples or spectral detail discarded by the MP3 encoder. A larger FLAC file is therefore not evidence of improved fidelity.
Similarly, FLAC to WAV does not increase quality: both can represent the same decoded PCM values. WAV may be convenient for editing compatibility; FLAC may save storage. Choose based on workflow, not a restoration claim.
File-size comparison formula
PCM bytes equal durationSeconds × sampleRate × bitDepth × channels ÷ 8, plus a small container header. The FLAC range multiplies that PCM estimate by the selected lower and upper ratios. Lossy bytes use durationSeconds × averageBitrateKbps × 1000 ÷ 8. Metadata and container overhead remain outside the simple payload numbers.
Because the FLAC ratio is explicitly a range, users can see that content matters. Noise-like material may compress less than predictable speech or tonal passages. Only encoding the actual source produces an actual FLAC size.
Keeping a master and making delivery copies
| Stage | Preferred source | Reason |
|---|
| Recording / capture | Original PCM or device master | Preserve the best available acquisition |
| Active editing | WAV or lossless project media | Avoid unnecessary lossy generations |
| Long-term archive | Documented lossless master | Preserve current sample values and context |
| Sharing / website / podcast | Tested lossy copy | Reduce transfer and storage for delivery |
A lossy delivery copy can be entirely appropriate. The important boundary is keeping it distinct from the best available master.