Audio Compressor

About Audio Compressor

This online audio compressor reduces the dynamic range of your audio, making loud parts quieter and quiet parts relatively louder. This creates more consistent, even audio that's easier to listen to and better suited for broadcasting, podcasting, and online content. Choose from professional presets or fine-tune advanced parameters for complete control.

What loudness target (LUFS) should I deliver to for each platform?

Most streaming services normalize to about -14 LUFS integrated: Spotify, Apple Music, Amazon Music and YouTube all sit around -14 LUFS, so mastering louder than that just gets turned down and you lose dynamics for nothing. Podcasts and spoken word target -16 LUFS (the Apple Podcasts and AES streaming-podcast recommendation), which keeps voices consistent and intelligible. Broadcast TV and radio follow EBU R128 at -23 LUFS integrated with a -1 dBTP true-peak ceiling (the US ATSC A/85 equivalent is -24 LKFS). Pick the matching target in the Loudness Target selector and the tool measures your compressed output and flags PASS or FAIL against it. If it fails, the one-click Normalize button re-runs the chain with loudnorm to land on the target.

What is the difference between a compressor and a true-peak limiter?

A compressor gently reduces gain above a threshold by a ratio (say 4:1), shaping the overall dynamics and feel of a track over many milliseconds via attack and release. A true-peak limiter is an extreme, fast compressor (effectively an infinite ratio) whose only job is to stop the signal from ever exceeding a ceiling such as -1 dBTP, catching inter-sample peaks that a normal sample peak meter misses. You compress to control musical dynamics and even out a performance; you limit to guarantee no clipping on the final master. This tool is a dynamic-range compressor; its loudness check reports true peak (dBTP) so you can see whether a real limiter is still needed before delivery, and the Normalize step applies loudnorm with a -1 dBTP true-peak guard.

How do I choose threshold versus makeup gain to hit a target loudness like -14 LUFS?

Think of it as two separate jobs. Threshold and ratio decide how much you tame the peaks: lower the threshold (or raise the ratio) until the gain-reduction tames the loudest moments without sounding squashed, usually a few dB of reduction on the peaks for music and a bit more for voice. Makeup gain then raises the whole, now-denser signal back up toward your target loudness. Because compression reduces crest factor, the same makeup gain buys you more perceived loudness than it would on an uncompressed file. The reliable workflow: compress first, then enable the Loudness Target, read the measured LUFS, and either nudge makeup gain or just click Normalize to let loudnorm hit -14 (or -16 / -23) LUFS precisely while keeping true peak under -1 dBTP.

What is LUFS, true peak (dBTP) and loudness range (LRA)?

LUFS (Loudness Units relative to Full Scale) is the standardized, ear-weighted measure of perceived loudness defined by ITU-R BS.1770 and EBU R128 — integrated LUFS is the single number platforms normalize to. True peak (dBTP) measures the actual analog peak after digital-to-analog reconstruction, including inter-sample overshoots that ordinary peak meters miss; keeping it at or below -1 dBTP prevents downstream clipping in lossy encoders. Loudness range (LRA) describes how much the loudness varies across the whole file in loudness units: a low LRA (a few LU) means very even, controlled dynamics, while a high LRA means wide swings. After compression this tool measures all three with FFmpeg's ebur128 filter so you can verify the delivery is compliant.

What is parallel compression and when should I use it?

Parallel (or New York) compression mixes a heavily compressed copy of a signal back underneath the original, uncompressed signal. The compressed layer raises the quiet detail and adds density and punch, while the dry layer preserves the natural transients and dynamics — you get loudness and consistency without the lifeless, squashed sound of heavy compression on its own. It is a staple on drum buses, mix buses and vocals. To approximate it here, compress fairly hard with a low threshold and high ratio for body, then back off makeup gain so the result blends with the source feel; for mastering, prefer gentler bus compression plus the loudness normalization step to reach your target LUFS rather than crushing a single pass.

How much gain reduction is too much, and how do I avoid over-compression?

Gain reduction is how many dB the compressor is pulling down at any moment; on a single compressor, 3 to 6 dB of reduction on peaks is plenty for most music and 6 to 10 dB suits spoken word that needs to be very even. Pumping, a loss of punch, dull lifeless transients and listener fatigue are the classic signs you have gone too far. Use a soft knee and moderate ratio (2:1 to 4:1) for transparency, reserve high ratios for genuine limiting, and let the release breathe with the tempo. Then confirm with the loudness readout: if you are already at -14 LUFS with a healthy loudness range, there is no reason to compress harder — reach for normalization instead of more ratio.