Audio Compressor
Compress audio files with format-aware optimization strategies for maximum efficiency
Audio Compressor
Compress audio files with format-aware optimization strategies for maximum efficiency
CHOOSE FILES
(or drag them here)
Understanding Audio File Sizes
Uncompressed (WAV, AIFF)
Uncompressed audio (WAV, AIFF) stores every single sample as a raw number. CD-quality audio uses 44,100 samples per second, each stored as a 16-bit value, in two channels (stereo). That is 44,100 x 2 bytes x 2 channels = 176,400 bytes per second, or about 10.6MB per minute. A 60-minute album in WAV format consumes roughly 640MB. This is the baseline that all compression works against.
Lossless (FLAC, ALAC, WavPack)
Lossless compression (FLAC, ALAC, WavPack) reduces file size by 30-60% without losing a single bit of audio data. FLAC achieves this through linear prediction: it guesses each sample based on previous samples, then stores only the prediction error. Music with lots of silence or repetition compresses better than chaotic noise. A classical recording with quiet passages might compress to 40% of WAV size, while heavy metal stays closer to 65%.
Lossy (MP3, AAC, Opus, Vorbis)
Lossy compression (MP3, AAC, Opus, Vorbis) achieves 80-95% reduction by permanently removing audio data that human ears are unlikely to notice. The psychoacoustic model identifies masking effects: a loud cymbal crash masks a quiet guitar note happening at the same time, so the encoder discards the quiet note. Frequencies above 16kHz are cut at lower bitrates because most adults cannot hear them. These decisions are irreversible.
Audio Format and Bitrate Comparison
| Format | Type | Size per Minute | Quality | Compatibility |
|---|---|---|---|---|
| WAV 16-bit/44.1kHz | Uncompressed | 10.6 MB | Perfect (reference) | Universal |
| FLAC | Lossless | 4-7 MB | Perfect (bit-identical to WAV) | Most players, no iOS native |
| MP3 320kbps | Lossy | 2.4 MB | Transparent for most listeners | Plays on everything |
| MP3 128kbps | Lossy | 0.96 MB | Good for speech, acceptable for music | Plays on everything |
| AAC 128kbps | Lossy | 0.96 MB | Better than MP3 at same bitrate | All modern devices |
| Opus 64kbps | Lossy | 0.48 MB | Excellent for voice, good for music | Browsers, Android, limited iOS |
Sizes are for stereo 44.1kHz content. Mono halves the size. Higher sample rates increase proportionally.
The Psychoacoustic Model Behind Lossy Compression
MP3 and AAC encoders do not simply throw away random audio data. They use a mathematical model of human hearing to decide what to keep and what to discard. The model is built on decades of research into auditory perception, starting with the work of Ernst Zwicker in the 1960s and refined through the MPEG standardization process.
Simultaneous Masking
When a loud sound and a quiet sound occur at the same time and at nearby frequencies, the loud sound prevents your ear from hearing the quiet one. The encoder measures the masking threshold for each frequency band in each time window and discards everything below that threshold. A loud bass drum at 100Hz masks quiet harmonics at 120Hz, so those harmonics use zero bits.
Temporal Masking
Temporal masking exploits the fact that a loud sound briefly deafens you to quiet sounds that come immediately before or after it. Pre-masking lasts about 5 milliseconds. Post-masking lasts 50-100 milliseconds. The encoder can allocate fewer bits to the moments right after a loud transient (like a snare hit) because your ears are temporarily less sensitive.
Absolute Hearing Threshold
The absolute threshold of hearing varies by frequency. Human ears are most sensitive between 1-5 kHz (the range of human speech). Below 100Hz and above 14kHz, sensitivity drops dramatically. Encoders exploit this by using fewer bits for very low and very high frequencies, especially at lower bitrates. This is why low-bitrate MP3s sound "muffled": the encoder aggressively cuts high-frequency content to save bits for the perceptually critical midrange.
Choosing the Right Settings for Your Use Case
Podcast Distribution
Mono MP3 at 96kbps. Podcasts are speech, which has a narrow frequency range. Mono is fine because spatial positioning does not matter for a single voice. 96kbps mono MP3 produces about 0.72MB per minute. A 45-minute episode is 32MB. Apple Podcasts, Spotify, and RSS readers all accept this format.
Music for Streaming
Stereo AAC at 192-256kbps. AAC handles musical transients and stereo width better than MP3 at equivalent bitrates. Spotify uses Ogg Vorbis at 320kbps for premium; AAC at 256kbps is perceptually equivalent. Apple Music uses AAC at 256kbps for its entire library.
Voice Memos and Dictation
Mono AAC at 64kbps or Opus at 32kbps. Voice has limited bandwidth (roughly 300Hz-3.4kHz for telephony, up to 8kHz for wideband). Ultra-low bitrates work well because the psychoacoustic model has very little to preserve. A 1-hour meeting recording at 64kbps AAC mono is about 29MB.
Archiving Vinyl Rips
FLAC is the correct choice for archival. If disk space forces lossy compression, use AAC or Opus at 256kbps minimum. Never use MP3 for archival: the format is 30 years old and less efficient than modern codecs. Once you compress to lossy, the discarded data is gone forever.
Frequently Asked Questions
Does compressing audio reduce quality?
It depends on the compression type. Lossless compression (FLAC, ALAC) reduces file size without any quality loss. Lossy compression (MP3, AAC, Opus) permanently removes audio data but at 192kbps or higher, the difference is inaudible to most listeners in blind tests.
What is the best free audio compressor online?
utilAZ is one of the best free audio compressors online. It supports MP3, WAV, M4A, AAC, FLAC, and OGG input formats with adjustable bitrate settings, server-side processing, no file limits, and no signup required.
How much can you compress an audio file?
Uncompressed WAV to MP3 at 128kbps yields about 90% reduction. A 50MB WAV becomes roughly 5MB. Already-compressed MP3 at 320kbps re-encoded to 128kbps drops to about 40% of original size. FLAC to MP3 typically gives 70-80% reduction.
How to compress audio for WhatsApp?
Upload your audio to utilAZ and compress to MP3 at 128kbps. WhatsApp accepts audio files up to 16MB. A 5-minute voice recording at 128kbps MP3 is only about 5MB, well within the limit while maintaining clear voice quality.
What is the best format for compressed audio?
For maximum compatibility, MP3 plays on every device ever made. For better quality at lower bitrates, AAC is superior (used by Apple Music, YouTube). For the best compression efficiency, Opus delivers excellent quality at very low bitrates but has limited device support.