Dynamic range analysis: DR, crest factor & LRA

Dynamic range analysis measures how much your levels move — from the quiet passages to the peaks. Too little and the mix is fatiguing and lifeless; too much and it feels inconsistent. Unveil reports dynamic range, crest factor (peak-to-RMS punch), and EBU R128 loudness range (LRA), then scores them against what your genre actually delivers.

What dynamic range is and why it matters

Dynamic range describes the distance between the loud and quiet parts of a recording. In music mastering it is usually discussed through several related measurements, because no single number captures everything. A track can have large dynamics between sections but very little within each hit, or the reverse.

It matters because dynamics carry emotion and impact. A chorus feels big partly because the verse before it was smaller. A kick feels powerful partly because it rises sharply above the level around it. When these relationships are flattened, the music sounds loud but flat, and listeners tire sooner.

The measurements: crest factor, PLR, DR and LRA

• Crest factor is the ratio of peak level to RMS level, in dB. A pure sine has a crest factor of about 3 dB. Unprocessed music often measures 12 to 20 dB, and heavily limited masters can fall to 6 to 8 dB. It describes short-term punch.
• Peak-to-loudness ratio (PLR) is the true peak minus the integrated loudness in LU. A master at -1 dBTP and -9 LUFS has a PLR of 8. It is a single-number summary of how much headroom exists above the average level across the whole track.
• DR value usually refers to a histogram-style dynamic range reading that compares peaks and loud-section RMS over the track. Different tools implement it differently, so compare values from the same tool only.
• Loudness range (LRA) is defined by EBU Tech 3342 and measures the spread of short-term loudness over the programme, between the 10th and 95th percentiles, after gating. It describes macro dynamics, how much louder the loud sections are than the quiet ones, and is expressed in LU.

How compression and limiting change each number

Compression and limiting act on different measurements. A compressor with a slow attack lets transients through and mostly reduces the body of loud sections. Crest factor can hold steady or rise, while LRA shrinks. A fast-attack compressor or a limiter shaves peaks, which lowers crest factor and PLR while raising the average.

Raising the level into a limiter is the most common cause of falling PLR. If you add 4 dB of gain into a limiter, the average goes up roughly 4 dB, the peaks stay stuck at the ceiling, and PLR falls by roughly 4 dB. That is the entire mechanism behind loudness war masters.

Clipping affects crest factor sharply because it cuts peaks without a time constant, but it also adds harmonic distortion. Automation or manual rides, which change level between sections, affect LRA but not crest factor.

Reference ranges by genre

These are common observations for commercial masters, not requirements. Individual releases deviate in both directions.

• Pop, EDM and hip-hop masters: PLR often 7 to 10 dB, LRA often 3 to 6 LU.
• Rock and indie: PLR often 8 to 12 dB, LRA often 5 to 9 LU.
• Acoustic, folk and singer-songwriter: PLR often 12 to 18 dB, LRA often 6 to 12 LU.
• Jazz and classical: PLR often 15 dB and above, LRA often 8 to 20 LU or more, depending on the repertoire.
• Pre-master mixes: aim to preserve dynamics, with no limiter on the bus and drum crest factors comfortably above 10 dB.

Over-compression: symptoms and fixes

Listen for these signs, then confirm with the numbers: drums lose their attack and sound like they are behind the vocals; the chorus does not feel bigger than the verse; sustained notes seem to breathe or pump with the kick; and the track feels tiring after a minute at moderate volume. Numerically, you will often see PLR below about 6 dB, crest factor of the full mix under 8 dB, and LRA of 2 LU or less on a song that is musically dynamic.

Fixes belong at the right level of the chain.

• Reduce gain into the final limiter. A smaller push is the most direct way to restore PLR.
• Lighten bus compression. Try 1 to 2 dB of gain reduction rather than 4 to 6, or use a slower attack to let transients through.
• Look for stacked processors. Three compressors each doing 3 dB of reduction add up to the same squash as one doing 9.
• Use clipping on individual drums rather than a final limiter to control peaks.
• Restore section contrast by automating levels, with the chorus 1 to 2 dB above the verse, instead of compressing it flat.
• Compare with the pre-master. If the unmastered mix has a healthy crest factor and the master does not, the damage was done in mastering.

Step-by-step analysis workflow

1. Measure integrated loudness and true peak, then compute PLR.
2. Read the LRA across the whole track.
3. Check the short-term loudness graph. Verses and choruses should be visibly different.
4. Measure crest factor on the drum bus or on a representative loud section.
5. Compare with two or three reference masters in the same genre, measured the same way.
6. Decide which number is out of line and trace it to the process responsible.
7. Change one thing, bounce, and measure again.

Worked example and checklist

A rock mix has -1.0 dBTP and -7.5 LUFS, so PLR is 6.5 dB. Its LRA is 2.8 LU. For the genre, both are low. Looking at the chain, the bus compressor does 5 dB of reduction and the limiter does 6 dB more. Reducing the bus compressor to 2 dB and the limiter to 3 dB, while keeping the -1.0 dBTP ceiling, gives integrated loudness of about -10.5 LUFS and a PLR of 9.5 dB, with LRA rising to about 4.5 LU. On a -14 LUFS platform the new version is turned down by 3.5 dB instead of 6.5 dB, so the end result is a similar playback loudness with more punch retained.

• PLR and LRA are in a plausible range for the genre.
• Short-term loudness shows clear section contrast.
• Drum crest factor has not collapsed.
• No more than one or two stages of heavy gain reduction in the chain.
• Compared against references at matched loudness.

What you get

• DR, crest factor, and LRA — what each one tells you and how they differ.
• Genre-aware targets: an EDM master and an acoustic record should not have the same dynamics.
• How to spot over-compression (squashed transients) before it's locked into the mix.
• A free dynamics readout with the moment your dynamics are worst.

Questions

Related

• Frequency spectrum analysis, explained for producers
• Stereo field analysis: width, phase & mono safety
• Free online LUFS checker
• Mix analysis guides
• Logic Pro mix analysis