If you just want the answer: use FLAC (or WAV) to keep a master copy of music you care about, AAC or MP3 at 192 to 256 kbps for everyday listening, and 64 to 96 kbps mono for speech. MP3 is the safest choice when compatibility matters most. AAC gives slightly better quality at the same bitrate. FLAC is the only one of the three that preserves the recording exactly.
The rest of this guide explains why, so you can make the call yourself instead of memorising numbers. When you are ready to convert, our Audio Converter handles all three formats.
Lossy vs lossless: the one idea that matters
Every audio format falls into one of two camps.
Lossless formats (FLAC, WAV, AIFF) keep the audio data complete. WAV and AIFF store it uncompressed; FLAC compresses it the way a ZIP file would, so you can decompress back to the identical samples. Nothing is lost, but files are large.
Lossy formats (MP3, AAC, Ogg Vorbis, Opus) use psychoacoustic models to throw away sound that you are least likely to notice, such as very quiet details masked by louder ones. The result is a much smaller file that cannot be restored to the original.
Uncompressed CD-quality audio (44.1 kHz, 16-bit, stereo) runs at 1,411 kbps. That is roughly 10.6 MB for every minute. A 128 kbps MP3 of the same minute is about 1 MB, and a 320 kbps MP3 is about 2.4 MB. FLAC typically lands somewhere around half the size of the WAV, though the exact ratio depends on the music; quiet, sparse recordings compress better than dense, loud ones.
For a broader look at this trade-off across images, video and documents, see lossy vs lossless file conversion explained.
The three formats side by side
The size column is an estimate derived from bitrate arithmetic, not a measurement of any particular song.
MP3
MP3 dates from the early 1990s and is still the lowest-risk option. Every phone, car, smart speaker, browser and editing program plays it. The cost is age: its encoders are mature but the format is less efficient than newer ones, so you need a higher bitrate to reach the same quality as AAC.
AAC
AAC was designed as MP3's successor. At equal bitrates it generally sounds a little cleaner, particularly below 192 kbps, which is one reason Apple and YouTube use it. You will meet it as .m4a files (AAC in an MPEG-4 container) or as raw .aac. Support is near-universal on modern devices, but very old players occasionally choke on it.
FLAC
FLAC is the open lossless standard. It supports tags and cover art, handles high-resolution audio, and is smaller than WAV. Use it where fidelity matters: ripping your CD collection, keeping studio mixes, or storing recordings you may edit later. Current iOS, Android, macOS, Windows and Linux can all play it, but check older hardware before relying on it.
Where WAV, Ogg and Opus fit
- WAV/AIFF: uncompressed, ideal as an editing format, wasteful for storage.
- Ogg Vorbis: an open lossy format, common in games and some open-source players.
- Opus: excellent at low bitrates, especially for voice, and widely used in calls and messaging. Support in dedicated music players is patchier than for MP3 or AAC.
Choosing a bitrate: 128, 192, 256, 320 and VBR
Bitrate is how many kilobits of data represent each second of audio. More bits means more detail retained and a bigger file. The gains shrink as you go up.
Sizes are calculated from bitrate (kbps multiplied by 60, divided by 8,000) and ignore small container overhead.
Constant vs variable bitrate (VBR)
A constant bitrate (CBR) file spends the same number of bits on every second, whether it is silence or a full orchestra. Variable bitrate (VBR) lets the encoder spend more bits on complex passages and fewer on simple ones, aiming for consistent quality instead of consistent size.
For music, VBR at a high quality setting is usually the better trade: you get smaller files for the same perceived quality. CBR is still handy when a platform or device needs predictable file sizes or streaming behaviour. Either way, a higher-quality VBR setting will beat a low-bitrate CBR file.
One honest caveat: your ears, headphones and the music itself decide what is audible. A blind test on your own equipment tells you more than any chart. If you cannot hear the difference between 192 and 320 kbps, the smaller file is the right one for you.
Which format for which job
Podcasts and spoken word
Speech has a narrow frequency range, so it compresses well. A mono file at 64 to 96 kbps in MP3 or AAC sounds clean for a single voice, and cuts file size to a fraction of a stereo music-quality file. Use 128 kbps if your show has music beds or you want extra margin. Check your hosting platform's guidelines first, since some re-encode what you upload.
Music you want to keep
Archive in FLAC (or the original WAV). From that master you can always make smaller copies for phones or sharing. Going the other way is impossible: you cannot rebuild a lossless file from an MP3.
Everyday listening on a phone
AAC or MP3 at 192 to 256 kbps gives you plenty of quality and lets thousands of songs fit in storage. If you listen mostly through a Bluetooth speaker or earbuds, higher bitrates buy you very little because the wireless link itself compresses audio again.
Voice notes and recordings
Voice memos are often long. Compress to 64 kbps mono to save space while keeping the speech perfectly intelligible. If the recording is evidence or an interview you might transcribe carefully, keep the original too.
Ringtones
Ringtones are short, so file size barely matters. What does matter is the format. iPhones need M4R (AAC in an MP4 container) at 40 seconds or less; Android phones typically accept MP3, M4A or Ogg and let you pick a file from storage. FileCrank's Audio to Ringtone tool trims a clip using start and end times and exports an M4R. In our implementation it encodes the clip as AAC at 128 kbps and clamps it to 40 seconds. You still have to add the file to your iPhone yourself, for example through Finder or iTunes, and you should only use audio you have the rights to use.
How to convert audio without wrecking quality
Follow these rules in order of importance.
- Start from the best source. If you have a FLAC or WAV, convert from that. Converting an MP3 to AAC means encoding twice, and the second pass compounds the damage.
- Convert once. Do not run a file through several lossy conversions. If you need several versions, make each one from the master.
- Do not expect upscaling to help. Converting a 128 kbps MP3 to 320 kbps or to FLAC makes the file larger but not better.
- Choose the format for the destination. MP3 for maximum compatibility, M4A/AAC for Apple-centric use, FLAC for archives.
- Keep metadata. Check that tags such as title and artist survive; our converter copies them across from the source.
Converting with FileCrank
- Open the Audio Converter and upload your file.
- Choose the output format (MP3, AAC, M4A, FLAC, WAV, OGG, AIFF or Opus).
- Run the conversion and download the result.
Some honest notes on how it behaves today. The tool picks sensible encoder settings per format rather than exposing every knob: MP3 output uses a high-quality variable-bitrate mode, AAC and M4A use 192 kbps, Opus uses 128 kbps, and WAV and AIFF are written as 16-bit PCM. Each tool page lists the current file size limit, and the limit differs between tools, so check it before you upload a long recording. Uploaded files are deleted automatically after about an hour.
For a wider tour of what is available, see our overview of free online audio tools.
How to compress audio sensibly
Compressing audio means re-encoding it at a lower bitrate. There is no free lunch, so do it with intent.
A simple decision process
- Decide the target. Email attachment limit? Podcast hosting quota? Phone storage? Work backwards from that size.
- Match the bitrate to the content. Speech can go to 64 kbps. Music should usually stay at 128 kbps or above, with 192 kbps a comfortable floor if you care about sound.
- Use mono for single-voice content. Dropping the second channel halves the data with no loss for a one-person recording. Skip this for stereo music or interviews recorded with spatial separation.
- Test a short sample. Compress a minute, listen on the speakers or headphones your audience will use, then do the full file.
- Keep the original. Compression is one-way.
Using the FileCrank compressor
The Audio Compressor re-encodes your file as an MP3 at one of three quality levels: Small file (64 kbps), Balanced (128 kbps) and High quality (192 kbps). Upload the file, pick a level, and download the result. Because the output is always MP3, it is a good fit when you want something that plays everywhere, but it is not the right tool if you need to stay in FLAC or AAC. Use the converter for that.
As a rough guide: pick Small file for voice notes and talk recordings, Balanced for casual music or podcasts with music, and High quality for music you actually plan to listen to carefully. If your source is already a low-bitrate MP3, compressing it again will only make it worse, so check the existing bitrate first.
Common mistakes to avoid
- Treating 320 kbps as a quality guarantee. A 320 kbps MP3 made from a 128 kbps source is still a 128 kbps recording in a larger wrapper.
- Archiving in MP3. Fine for sharing; poor for a master, because every future conversion starts from damaged audio.
- Using stereo for everything. Spoken-word files rarely need it.
- Ignoring the destination. Some platforms re-encode anything you upload, so uploading a huge lossless file may not improve what listeners hear.
- Forgetting rights. Converting your own recordings is fine. Using other people's music for ringtones, videos or redistribution may need permission.
Quick reference checklist
- Master copy of music: FLAC or WAV
- Compatible sharing: MP3, 192 to 256 kbps
- Apple or streaming workflow: AAC/M4A, 192 to 256 kbps
- Podcast voice: 64 to 96 kbps, mono
- Voice memos: 64 kbps, mono
- iPhone ringtone: M4R, 40 seconds maximum
- Never chain lossy conversions; always go back to the original
Conclusion
The best audio format is the one that matches what you will do with the file. Keep lossless masters of anything irreplaceable, hand out lossy copies at a bitrate your ears and storage agree on, and resist converting upwards or repeatedly. If you want to try it, convert your audio, shrink a large recording, or turn a favourite clip into a ringtone. No account is needed.