Why audio sounds too quiet
A recording may be quiet because its peaks are low, its average level is low, the performance has a wide dynamic range, or the playback platform applies loudness normalization. A file can peak near 0 dBFS and still sound quiet if only a few transients reach that level. Conversely, a dense recording can sound loud with lower peaks. Measure peak, True Peak and LUFS before changing anything; one number alone does not explain perceived loudness.
Gain and peak normalization
Gain raises every sample by the same amount. Peak normalization calculates the gain needed to place the highest peak at a target such as -1 dBFS. It is transparent because it does not change internal dynamics. However, it may not make a very dynamic file much louder: one isolated peak can limit the gain for the entire recording. Peak normalization is ideal when the balance is already correct and the file simply has unused headroom.
Loudness normalization
Loudness normalization targets an integrated LUFS value rather than the highest sample. It is useful for podcasts, videos and playlists that need consistent perceived volume. If the required increase would push peaks too high, gain alone is insufficient. A limiter may be needed, but large reductions can sound aggressive. Platform targets are not mastering laws; they are playback references. Preserve musical dynamics and leave safe True Peak headroom for encoding.
Compression before limiting
Compression reduces the difference between louder and quieter passages. With appropriate threshold, ratio, attack and release, it can make speech steadier or bring low-level detail forward. Too much compression causes pumping, dull attacks and listener fatigue. Use makeup gain only after judging the gain reduction. For natural results, aim for controlled peaks rather than a constantly moving meter. Parallel compression can add density while preserving some original transients.
Using a limiter safely
A limiter prevents peaks from crossing a ceiling and allows additional level. Set an output ceiling below 0 dBFS, often with True Peak protection for compressed delivery. Watch how much gain reduction occurs on the loudest sections. Occasional small reductions may be transparent; continuous heavy limiting flattens transients and can increase distortion. Compare versions at matched playback level so “louder” does not automatically seem better.
Recommended workflow
Analyze the source and identify whether the issue is unused headroom or excessive dynamics. For a balanced file, use peak or loudness normalization. For uneven speech or music, apply gentle compression, then a limiter only if necessary. Check Integrated LUFS and True Peak after processing and after export. Listen for clipped consonants, flattened drums and pumping ambience. Keep a lossless master and use WaveFixer to normalize a delivery copy in the browser.