What's the difference between lossless and lossy?

Lossy throws away part of the sound forever to shrink the file — that's how MP3 and AAC work. Lossless compresses audio so it can be restored byte-for-byte — that's FLAC and ALAC. Lossless is identical to the original; lossy is just a very close copy.

Can you hear the difference between FLAC and MP3 320?

Usually no. In blind ABX tests most people can't reliably tell FLAC from a 320 kbps MP3 on typical gear. Catching a difference is rare and only happens with good equipment, familiar tracks and a quiet room.

ABX is a blind test: you're given samples A and B plus a mystery X, and you must decide whether X matches A or B. If you're right no more often than a coin flip, you can't hear the difference. Around 20 trials are usually run so the result is statistically significant.

Do I need Hi-Res audio at 24-bit / 96 kHz?

For listening, essentially no: in blind tests people can't tell Hi-Res from standard CD quality (16-bit / 44.1 kHz). The extra resolution helps during recording and mixing, but for a finished track it's mostly marketing. Standard FLAC from a CD is more than enough.

When is FLAC actually worth it?

When you're archiving a collection and want a lossless master; when you plan to re-convert your music into different formats later; when you're editing or mixing audio. And for peace of mind too: it's nice to know you're storing an exact original, not a copy of a copy.

Lossless vs Lossy Audio: Can You Really Hear the Difference

Audiophiles swear lossless sounds "livelier" and "wider," while skeptics reply, "you'll never tell them apart in a blind test." Who's right? We won't guess — we'll look at how hearing actually works and at the results of honest ABX tests. And along the way, decide whether your collection is worth the megabytes of FLAC.

The fundamental difference

It all comes down to one thing: what happens to the data during compression.

Lossy — formats like MP3, AAC and Opus. They throw away part of the audio data forever to make the file small. You can't restore the original: it's like retelling a book in your own words — the meaning survives, but the exact text is gone.

Lossless — formats like FLAC, ALAC and WavPack. They also shrink the file, but in a clever way from which the original audio can be restored byte-for-byte. It's like zipping a document: the file is smaller, yet unpacking gives you an exact copy. That's why FLAC is mathematically identical to the source WAV — zero loss.

The key point

"Lossless is identical to the original" is a fact about the data, not about what you'll hear. The audible difference between good lossy and lossless is a separate question — and a far less obvious one.

How perceptual coding works

"Throwing away part of the sound" might sound crude and obviously audible. But lossy codecs are subtler. At their core is perceptual coding: they discard not random data but exactly the parts the human ear doesn't perceive anyway. Psychoacoustics knows several such "blind spots":

Loud masking quiet. If a quiet sound follows right after a loud one, the ear simply doesn't notice it — so the codec removes it freely.
The limits of hearing. Most adults can't hear above 16-18 kHz. Anything higher can be cut with almost no consequences.
Simultaneous masking. A loud tone "drowns out" quiet sounds near it in frequency — those can be skipped too.

That's why a 320 kbps MP3 sounds so close to the original: it mostly discarded things you wouldn't hear anyway. Trouble starts at low bitrate, where the codec has to cut into the audible — and that's when the familiar artifacts appear.

Close-up of pink wireless headphones on a wooden table in warm red light — On good headphones in silence the difference is easiest to catch — yet even there it's barely perceptible. Photo: Pexels

The ABX test: check your own ears

Claims of "I can hear the difference" are worthless until they survive a blind test. The gold standard here is ABX. It works like this: you're given reference A (say, FLAC) and reference B (the same track as MP3), then a mystery sample X that randomly equals A or B. Your job is to decide which one X is. Over and over.

The trick is in the statistics. If you're just guessing, on average you'll be right half the time — like a coin flip. To prove you really hear a difference, you need to be right noticeably more than 50%. That's why around 20 trials are used: hitting 16-17 out of 20 by luck is nearly impossible, but if you genuinely hear it, you'll get there.

Try it yourself

There are free online ABX tests (for example at abx.digitalfeed.net): upload your track, listen blind, and get honest statistics. It's the fastest way to learn whether you personally need lossless — with no one else's opinion involved.

What the tests show in practice

The results of dozens of amateur and scientific ABX tests converge on one conclusion: at 320 kbps most people can't reliably tell MP3 from lossless. A typical story is a participant scoring 16 out of 25 correct — not enough for statistical significance: they might as well have been guessing. Many honestly admit they caught no difference at all when tested blind.

This doesn't mean there's no difference for anyone. An experienced listener on top-tier gear, on certain tracks that are "hard" for the codec, in silence, sometimes really does hear it. But that's the exception, not the rule, and it applies to high bitrates. At 128 kbps the gap with lossless is heard far more often — which is exactly why bitrate matters more than the mere "lossy vs lossless" label.

~50%accuracy = no audible difference

≈20trials for a significant test

320kbps where lossy is near-transparent

When FLAC actually matters

If the difference is barely audible, why bother with lossless at all? It turns out there are solid reasons — they're just not always about "sounds better right now":

Archive and master copy. FLAC keeps a perfect original. From it you can make an MP3 at any bitrate anytime — but not the other way around.
Future conversions. When a new, more efficient codec appears, you'll re-encode your collection from FLAC without loss. From MP3 you can't: the loss keeps piling up.
Editing and processing. Cutting, mixing and applying effects should be done on lossless material, or artifacts creep in.
Peace of mind. Yes, that counts too: it's nice to know you're holding an exact original, not a "copy of a copy."

The difference between lossless and lossy is less about what you hear today than about what you'll be able to do with the sound tomorrow.

Hi-Res 24/96: truth or marketing

A separate trendy topic is Hi-Res audio: 24-bit and 96 kHz instead of the CD-standard 16-bit / 44.1 kHz. Sellers promise "studio quality" and charge a premium for it. But here's the inconvenient truth: in blind tests people can't tell Hi-Res from ordinary CD quality when listening.

The reason is simple: 16-bit and 44.1 kHz already cover the full range of human hearing with room to spare. The extra resolution really helps during recording and mixing — there's headroom to cut and process. But in a finished track you merely listen to, the extra bits go nowhere. So standard FLAC from a CD is more than enough, and paying extra for Hi-Res only makes sense if it matters to you on principle.

The verdict

Let's put it together, short and to the point:

Just listening to music

MP3/AAC at 256-320 kbps is transparent. Lossless brings no audible gain.

Building a collection for the long haul

Store it in FLAC. It's quality insurance and freedom to convert later.

Want to test your ears

Run an ABX test. Honest statistics settle the argument better than any opinion.

Need to change a file's format?

FormatZ converts images, documents and other files online — no install, no sign-up, full privacy. Everything runs right in your browser.

Open the FormatZ converter

Want to cement the topic? Get to grips with what audio bitrate is — it affects audibility more than the lossy/lossless label itself. Deciding on a storage format? The MP3 vs WAV vs FLAC comparison helps. And if you've decided to move your library to FLAC, read how to convert music without loss.