How to Compress Audio Files Without Wrecking the Sound

Audio files have a way of ballooning out of control. A three-minute song ripped to WAVcan easily top 30 MB, a one-hour recorded interview can run past a gigabyte, and email and chat apps start rejecting attachments long before that. The good news is that audio compresses extremely well — if you understand a few core settings. Get them right and you can shrink a file by 80–90% with no difference your ears can detect. Get them wrong and you end up with tinny, swirly, underwater-sounding audio. This guide walks through the choices that actually matter.

Why WAV and AIFF files are so big

WAV (and its Apple cousin AIFF) stores audio as raw, uncompressedPCM samples. Nothing is thrown away and nothing is packed down, so the size is entirely predictable from three numbers: the sample rate, the bit depth, and the number of channels. CD quality — 44,100 samples per second, 16 bits per sample, two channels — works out to about 10 MB per minute. That is why a full album in WAV can fill several hundred megabytes while the same album as MP3 fits on a phone with room to spare.

Uncompressed audio is the right choice while you are recording and editing, because every edit and effect stays pristine. But it is a terrible choice for sharing or storing the finished result. The moment you are done editing, it is worth converting to a compressed format.

Lossy vs lossless: MP3/AAC vs FLAC

Audio compression comes in two flavors. Lossless formats like FLAC and ALAC shrink the file while keeping every original sample intact — decode a FLAC and you get back a byte-perfect copy of the source. They typically cut size by 40–60%, so that 10 MB-per-minute WAV becomes roughly 5–6 MB per minute. Use lossless when you are archiving masters or care about perfect fidelity.

Lossyformats like MP3 and AAC go much further by permanently discarding sound that the human ear is unlikely to notice — frequencies masked by louder ones, detail above your hearing range, and so on. This is how psychoacoustic encoding squeezes a song down to a tenth of its original size. The trade-off is that the discarded data is gone for good, so re-encoding a lossy file repeatedly slowly degrades it. If you want the full mental model, see our guide on lossy vs lossless compression.

Between the two modern lossy options, AAC generally sounds a little better than MP3at the same bitrate and is the default for Apple Music, YouTube, and most streaming. MP3 still wins on universal compatibility — if you need a file that plays on literally any device made in the last 25 years, MP3 is the safe pick.

The setting that matters most: bitrate

For lossy audio, bitrate— measured in kilobits per second (kbps) — is the single biggest lever on both quality and size. Higher bitrate means more data kept per second of sound, which means better quality and a larger file. The trick is matching the bitrate to the content.

For music, where you want full-range fidelity:

• 128 kbps— acceptable for casual listening on earbuds; the smallest you should go for music.
• 192 kbps— a great everyday sweet spot; most people cannot reliably tell it from the original.
• 256–320 kbps— near-transparent, the right choice if you have good headphones or a critical ear.

For speech— podcasts, voice memos, audiobooks, interviews — you can go much lower because the human voice occupies a narrow frequency range:

• 64 kbps— perfectly clear for a single voice; ideal for spoken-word podcasts.
• 96 kbps— a comfortable margin for speech with some background music or two speakers.

Encoding a one-hour podcast at 64 kbps instead of 192 kbps takes it from about 86 MB down to roughly 29 MB — a third of the size, with no meaningful loss in clarity for voice.

Sample rate, bit depth, and mono vs stereo

Three more settings can trim size further when used appropriately. Sample rate is how many times per second the audio is measured. 44.1 kHz (CD) and 48 kHz (video) capture the full range of human hearing, so there is rarely a reason to go higher for a final export. You can safely downsample a phone voice memo to 22.05 kHz, but leave music at 44.1 kHz or 48 kHz.

Channels matter too. A mono recording is half the size of stereo. If your source is a single narrator or a mono voice memo, exporting in mono is free savings with zero quality cost. Keep stereo for music and anything with a real left/right image.

For a thorough, browser-focused reference on which codecs do what and how they handle these parameters, the MDN guide to web audio codecs is the authoritative source.

Converting FLAC or WAV to MP3 or AAC for sharing

The most common real-world task is taking a big lossless file and turning it into something small enough to email, upload, or post. The recipe is simple:

• Start from your highest-quality source (the WAV or FLAC master).
• Choose AAC for best quality-per-byte, or MP3 for maximum compatibility.
• Pick a bitrate that fits the content: ~192 kbps for music, ~64–96 kbps for speech.
• Set mono for single-voice recordings; leave stereo for music.
• Keep the sample rate at 44.1 or 48 kHz for music; 22.05 kHz is fine for voice.

Because the conversion from lossless to lossy only happens once, you keep all the quality you can while still getting a dramatically smaller file. Just remember to hold on to the original master — never re-compress an already-lossy file when you can go back to the source.

Do it privately in your browser

You do not need to install anything or upload your recordings to a random server to compress them. FileShrinking’s audio compressorruns entirely in your browser, so your files never leave your device — nothing is ever uploaded. Drop in a WAV, FLAC, MP3, or AAC, choose your bitrate, and download the result. If your audio is actually part of a clip you are sharing, the same privacy-first approach applies to our video compressor. Experiment with a couple of bitrate settings and trust your own ears — since everything stays local, there is no risk in trying.