Audio Compare

About Audio Compare

This professional audio comparison tool lets you compare two audio files using AB testing, integrated LUFS loudness, true-peak (dBTP) and detailed spectral analysis. Perfect for loudness-matching masters, checking streaming compliance (-14 LUFS / -1 dBTP), and comparing codecs like MP3 vs FLAC or different bitrates. All processing happens locally in your browser for complete privacy.

What audio formats are supported?

All common audio formats are supported including MP3, WAV, OGG, AAC, M4A, FLAC, OPUS, and more. You can compare files in different formats.

What is AB testing?

AB testing is a method where you quickly switch between two audio sources (A and B) to directly compare their sound quality. This tool maintains the same playback position when switching, making it easy to hear subtle differences. Use the Switch button or play A/B individually.

What metrics are compared?

Integrated Loudness (LUFS), True Peak (dBTP), Sample Peak, RMS Level (average level), Spectral Centroid (brightness/tonal center), Spectral Rolloff (high-frequency content), Spectral Flatness (noisiness), and Zero Crossing Rate (percussiveness/transitions).

Which file is louder, and how do I A/B fairly?

Check the Integrated Loudness (LUFS) panel: the file with the higher LUFS is louder, and the 'Loudness Difference (LU)' tells you exactly how much to trim it. A louder file almost always 'sounds better' in a blind test, so for a fair A/B you should loudness-match first by lowering the louder source by that LU delta, then compare tone and detail.

What is LUFS and how is it measured here?

LUFS (Loudness Units Full Scale) is the standard loudness measurement defined by ITU-R BS.1770 and adopted by EBU R128. This tool computes integrated LUFS by applying the BS.1770 K-weighting filter (a high-shelf pre-filter plus an RLB high-pass), measuring 400 ms blocks with 75% overlap, then applying the -70 LUFS absolute gate and -10 LU relative gate before averaging — the same gating used by professional loudness meters.

What is True Peak (dBTP) and why does it differ from Sample Peak?

Sample Peak is just the maximum absolute sample value. True Peak (dBTP) reconstructs the analog waveform via 4x oversampling to catch inter-sample peaks that can exceed 0 dBFS after a DAC or lossy codec. A file can read -0.1 dB sample peak yet clip at +0.6 dBTP, so mastering and broadcast specs (e.g. -1 dBTP) are defined on true peak, not sample peak.

What LUFS should I target for Spotify, YouTube or Apple Music?

Most streaming platforms (Spotify, YouTube, Amazon, Tidal) normalize to about -14 LUFS integrated with a -1 dBTP ceiling. Apple Music targets around -16 LUFS. Broadcast under EBU R128 targets -23 LUFS. Pick a target in the dropdown and the panel shows a Compliant / Out of Spec verdict for each file so you can use this as a quick master QC check.

What is Spectral Centroid?

Spectral Centroid represents the 'center of mass' of the spectrum - essentially where most of the audio energy is concentrated. Higher values indicate brighter, more treble-focused sound. Lower values indicate darker, bass-heavy sound. Measured in Hz.

What is Spectral Rolloff?

Spectral Rolloff is the frequency below which a certain percentage (usually 85%) of the spectral energy is contained. It indicates the amount of high-frequency content. Higher values mean more high-frequency content, while lower values indicate more bass-focused audio.

What is Spectral Flatness?

Spectral Flatness measures how noise-like a sound is versus how tone-like it is. Values near 1.0 indicate white noise (flat spectrum), while values near 0.0 indicate pure tones (pitched sounds). Useful for distinguishing between musical content and noise.

What is Zero Crossing Rate?

Zero Crossing Rate (ZCR) is the rate at which the audio signal changes from positive to negative or vice versa. High ZCR indicates high-frequency content, noise, or percussive sounds. Low ZCR indicates low-frequency or sustained sounds. Useful for audio classification.

How do I use AB testing effectively?

1) Load both audio files, 2) Play Audio A and listen carefully, 3) Click Switch to instantly compare with Audio B at the same position, 4) Switch back and forth multiple times to hear differences, 5) Enable Loop to repeat a specific section. Focus on one aspect at a time (bass, treble, clarity, etc.).

What do the difference values mean?

For spectral metrics the difference shows how much Audio B differs from Audio A, with a percentage where meaningful. For logarithmic measurements (LUFS, True Peak, RMS, Sample Peak) we show only the absolute delta in dB or LU, because a percentage change of a dB figure is physically meaningless (e.g. -20 dB to -10 dB is not '+50%'). The loudness row reports the difference in LU.

Are the Hz values accurate across sample rates, and is analysis stereo?

Yes. Spectral Centroid and Rolloff are reported in true Hz, scaled to each file's own sample rate, so comparing a 48 kHz file against a 44.1 kHz file is apples-to-apples. Levels and loudness are measured on a full stereo-to-mono downmix (all channels summed), not just the left channel, so content panned anywhere in the stereo field is included. A Hanning window is applied before the FFT to reduce spectral leakage.

Can I compare different file formats?

Yes! You can compare any two audio formats. This is useful for testing lossy compression (MP3 vs FLAC), evaluating different bitrates (128kbps vs 320kbps), or comparing before/after audio processing.

What should I look for when comparing audio?

For loudness: Compare RMS and Peak levels. For brightness/tone: Check Spectral Centroid. For high-frequency detail: Look at Spectral Rolloff. For overall character: Compare Spectral Flatness. Use AB testing to hear these differences subjectively.

Is my audio file safe?

Absolutely! All audio analysis and playback happens directly in your browser using Web Audio API and Meyda. Your audio files are never uploaded to any server. Everything stays private on your device.