Guide
MP3, M4A, OPUS, WAV or FLAC? How to choose an audio format
A plain-English comparison of MP3, M4A, OPUS, WAV and FLAC: sound quality, file size, compatibility and which one to pick for music, podcasts and editing.
Lossy vs. lossless, in one minute
Digital audio starts as a long list of numbers called samples. CD-quality stereo stores 44,100 samples per second for each of two channels, at 16 bits per sample. That works out to 1,411.2 kilobits every second, or about 10.6 MB per minute. Every audio format is a different answer to the same question: how do you store those numbers in less space?
Lossy formats (MP3, AAC and Opus) use a model of human hearing to decide what you're unlikely to notice. Very quiet sounds that sit right next to loud ones, and detail masked by other frequencies, are simplified or dropped. The result is far smaller: a 192 kbps MP3 takes about one-seventh the space of CD audio, and at a sensible bitrate most people can't tell it from the original. What's discarded is gone for good.
Lossless formats keep every sample exactly. WAV stores them as they are. FLAC packs them more efficiently, the way a ZIP file packs a document, and unpacks them bit for bit when you press play.
A useful way to think about it: lossy is a great photograph of a painting, and lossless is a perfect copy of whatever you started with. The catch, as you'll see below, is what you start with.
The five formats up close
MP3: the one that plays everywhere
MPEG-1 Audio Layer III dates from the early 1990s, and its patents have since expired. Almost every device made in the last twenty-five years can play it, from phones and laptops to car stereos, smart speakers and old MP3 players. Its compression is less efficient than newer codecs, so it needs a slightly higher bitrate to sound just as good. At 192 kbps it's a safe, transparent-sounding choice for most listening.
- Type
- Lossy
- In Konvert
- 128, 192 or 320 kbps
- Size
- 0.96–2.4 MB per minute
- Plays on
- Practically everything
Not sure which bitrate to choose? The MP3 bitrate guide breaks down 128, 192 and 320 kbps with exact file sizes.
M4A (AAC): Apple's everyday format
AAC, or Advanced Audio Coding, was designed as MP3's successor and generally sounds better at the same bitrate. An .m4a file is simply AAC audio inside an MP4 container with no video track. It's the native format for iPhone, iPad, Mac and Apple Music, and Android, Windows and web browsers all play it too. Some very old players and car stereos only understand MP3.
- Type
- Lossy
- Size
- About 1 MB per minute at 128 kbps
- Plays on
- Apple devices, Android, Windows, browsers
OPUS: modern and efficient
Opus is an open, royalty-free codec standardized by the IETF in 2012 (RFC 6716). It blends a speech-focused engine with a music-focused one, so it handles both voices and full mixes well, even at low bitrates. That efficiency is why it powers web-based voice and video calls, and why YouTube typically streams its audio as Opus. Opus always works at 48 kHz internally. Support is excellent in browsers, Android and VLC, but patchier elsewhere. Apple's Music app, for example, won't import .opus files.
- Type
- Lossy
- Size
- About 0.7–1.2 MB per minute
- Plays on
- Browsers, Android, VLC, most media players
WAV: raw and ready to edit
A WAV file usually holds uncompressed PCM, the raw sample numbers with a small header in front. Nothing needs decoding, so every audio editor, video editor and digital audio workstation (DAW) opens it instantly and scrubs through it smoothly. The trade-off is size. At 16 bits and 48 kHz, stereo audio takes 48,000 × 2 channels × 2 bytes = 192,000 bytes a second. That's 11.52 MB per minute, or about 691 MB for an hour. At 44.1 kHz it's about 10.6 MB per minute.
- Type
- Uncompressed (lossless)
- Size
- About 10.6–11.5 MB per minute
- Plays on
- Everything; ideal for editors
FLAC: lossless, but smaller
The Free Lossless Audio Codec shrinks PCM audio without discarding anything. Decode a FLAC file and you get back exactly the samples that went in. For a typical 16-bit CD rip, a FLAC file is often around half to two-thirds the size of the WAV, depending on the music. It also supports proper tags, which WAV handles poorly. FLAC plays on Android, on Windows 10 and later, and in the Files app on iPhone. Apple's Music app uses its own lossless format (ALAC) and won't import FLAC.
- Type
- Lossless, compressed
- Size
- Varies with the audio; see below
- Plays on
- Android, Windows, VLC, audiophile players
Side-by-side comparison
| Format | Type | Size per minute | Compatibility | Best for |
|---|---|---|---|---|
| MP3 | Lossy | 0.96–2.4 MB | Universal | Music libraries, sharing, car stereos |
| M4A | Lossy | About 1 MB | Very wide | iPhone, iPad, Mac |
| OPUS | Lossy | 0.7–1.2 MB | Good, not universal | Speech, podcasts, saving space |
| WAV | Uncompressed | 10.6–11.5 MB | Universal | Video editing, DAWs, sampling |
| FLAC | Lossless | About 5–11.5 MB (varies) | Good, not on Apple Music | Archiving, editing without huge files |
Sizes are for stereo audio, with 1 MB counted as 1,000,000 bytes. FLAC sizes are for files made by Konvert from a lossy source.
Why the source matters most
This is the part that saves the most disk space and confusion. The audio in an online video is already lossy. YouTube, for example, typically delivers audio as Opus in a WebM file or AAC in an MP4 file, usually at bitrates well under 200 kbps. Whatever that encoder threw away is not in the video anymore, and no converter can bring it back.
So when you choose WAV, FLAC or a 320 kbps MP3, you're not recovering hidden quality. You're making a very faithful copy of an already compressed stream. A 320 kbps MP3 made from a 130 kbps Opus track doesn't sound like 320 kbps. It sounds like the Opus track, plus a tiny bit of new MP3 loss.
Higher settings avoid adding loss. They can't remove loss that already happened.
That doesn't make lossless formats pointless. Converting to WAV or FLAC means there's no second lossy generation. That matters when you're going to cut, process and export the audio again, because each lossy encode compounds the last. For editing, a lossless intermediate is the professional habit, and our guide to extracting audio for editing and podcasts goes deeper on that workflow.
What Konvert does with each format
Konvert uses yt-dlp to fetch the best audio-only stream a video offers, then FFmpeg to produce the format you picked. We checked exactly how that pipeline treats each choice:
- OPUS: if the source is already Opus, as it typically is on YouTube, the stream is copied into an .opus file without re-encoding. It's the closest you can get to the original audio. If the source is something else, it's encoded with FFmpeg's Opus encoder at its default setting (96 kbps for stereo).
- M4A: if the source is AAC, it's copied into an .m4a file with no re-encoding. Otherwise it's encoded to AAC at FFmpeg's default of about 128 kbps.
- MP3: encoded with the LAME encoder at the constant bitrate you choose: 128, 192, 320 kbps.
- WAV: decoded to 16-bit PCM.
- FLAC: decoded and compressed losslessly. When the source is Opus or AAC, FFmpeg keeps 24 bits of the decoded precision. Those extra bits are close to random and don't compress well, so these FLAC files are larger than a CD-rip rule of thumb suggests. In our tests they ranged from about half the size of the matching WAV to nearly the same size, depending on how dense the audio was.
Apart from OPUS, which always runs at 48 kHz, every format keeps the source's sample rate, so a file made from YouTube's Opus audio is 48 kHz. Files come without tags or cover art, named after the video's title. When you're ready, paste a YouTube link into the converter and choose the format that fits.
Which should you pick?
- Music library: MP3 at 192 kbps. It plays anywhere and sounds the same as the source to most ears. Choose 320 kbps if storage doesn't matter to you.
- iPhone and Apple Music: M4A. It imports into the Music app, plays natively and, from an AAC source, is copied without re-encoding.
- Podcasts and speech: OPUS for the smallest files, or MP3 at 128 kbps if the listening app needs MP3. Voices don't need high bitrates.
- Video editing or a DAW: WAV. It drops straight onto a timeline, and 48 kHz matches the usual video project setting. Pick FLAC if your editor supports it and space is tight.
- Archiving: OPUS to keep the original stream as-is in the least space, or FLAC if you want a lossless file that any future tool can convert cleanly.
- Transcription: M4A, MP3 or OPUS. Speech recognition tools don't benefit from lossless audio. Many, including OpenAI's Whisper, resample to 16 kHz mono internally anyway.
Questions about formats
Is OPUS better than MP3?
At the same bitrate, usually yes. Opus is a much newer codec and generally sounds cleaner than MP3 at low and medium bitrates, especially for speech. MP3 still wins on compatibility, so pick it when the file has to play on older players, car stereos or apps you don't control.
Does converting from one lossy format to another lose quality?
Yes, a little each time. Every lossy encode throws away some information, so turning an Opus or AAC stream into MP3 adds a second generation of loss. At 192 kbps or higher it's rarely audible, but if you plan to edit and re-export, choose WAV or FLAC to avoid stacking losses.
Why is my FLAC file almost as big as the WAV?
When the source is lossy audio such as Opus or AAC, FFmpeg decodes it to high-precision samples and keeps 24 bits of that precision in the FLAC file. The extra low-order bits are close to random, which FLAC can't compress well. The WAV is written at 16 bits, so the size gap is smaller than you'd see with a CD rip.
What sample rate will my file have?
MP3, M4A, WAV and FLAC files keep the sample rate of the audio Konvert downloads. Opus audio is always 48 kHz, so anything made from an Opus source is 48 kHz, and an OPUS file is 48 kHz whatever the source. AAC sources are usually 44.1 kHz or 48 kHz.