A practical guide to understanding audio metering, dynamics, and mastering decisions.
When mixing or mastering music, we’re confronted with a variety of measurements: LUFS, dBFS, True Peak, PSR, PLR, LRA, and more.
But what do these numbers actually tell us? And perhaps more importantly, which ones should we pay attention to?
The key is understanding that loudness, peak levels, and dynamics are different aspects of sound. A track can be loud but still punchy, dynamic but relatively quiet, or technically well-controlled yet musically lifeless.
No single measurement tells the whole story.
In this guide, I’ll explain the most relevant metering values, how to interpret them, and how I use them when mastering music.
1. LUFS – Perceived Loudness
LUFS stands for Loudness Units relative to Full Scale.
Unlike traditional peak meters, LUFS measures loudness in a way that better reflects human hearing.
This makes it particularly useful for comparing the perceived loudness of different tracks.
There are three main LUFS measurements.
Integrated Loudness (LUFS-I)
Integrated LUFS measures the average perceived loudness over an entire track or measurement period, using a gating system that excludes certain very quiet passages.
It is the primary reference for overall mastering loudness.
For example:
| Integrated LUFS | General character |
|---|---|
| −18 LUFS | Relatively quiet |
| −14 LUFS | Moderate loudness |
| −10 LUFS | Loud |
| −9 LUFS | Very loud |
| −7 LUFS | Extremely loud |
These descriptions are approximate. Two tracks with identical LUFS values can still sound quite different.
What to watch for: Use Integrated LUFS to establish your desired overall loudness, but don’t force every production toward the same number.
Short-Term Loudness (LUFS-S)
Short-Term LUFS measures perceived loudness over a moving three-second window.
It helps identify how loud individual sections of a track are.
For example, a beat could have:
- Integrated Loudness: −9 LUFS
- Verse: −11 LUFS Short-Term
- Chorus: −8 LUFS Short-Term
This tells us that the chorus is noticeably louder than the verse, even though the track’s overall loudness is −9 LUFS.
What to watch for: Compare verses, choruses, transitions, and climaxes. Short-Term LUFS is particularly useful when evaluating the balance between different song sections.
Momentary Loudness (LUFS-M)
Momentary LUFS measures loudness over approximately 400 milliseconds.
It reacts faster than Short-Term LUFS and reveals brief changes in perceived loudness.
It can be helpful when examining short musical events, but it should not be confused with the actual transient peaks of a kick or snare.
What to watch for: Momentary LUFS is useful for observing rapid loudness changes, but it’s usually less important than Integrated and Short-Term LUFS when mastering a complete track.
2. dBFS – Digital Peak Level
dBFS stands for decibels relative to Full Scale.
It describes signal amplitude in the digital domain, with 0 dBFS representing the maximum possible sample amplitude in conventional fixed-point PCM audio.
A sample-peak meter shows the highest individual digital sample values.
For example:
- −6 dBFS: 6 dB below full scale
- −3 dBFS: 3 dB below full scale
- −1 dBFS: 1 dB below full scale
- 0 dBFS: Maximum sample amplitude
A sample peak of −3 dBFS leaves 3 dB of sample-level headroom before reaching full scale.
This is particularly relevant when preparing audio for further mixing or processing.
What to watch for: Sample peaks are useful for gain staging, headroom management, and preparing unfinalized audio.
However, staying below 0 dBFS does not guarantee that the reconstructed audio waveform will remain below the same level.
That’s where True Peak comes in.
3. True Peak (dBTP) – Peaks Between the Samples
True Peak estimates the maximum amplitude of the reconstructed audio waveform, including peaks occurring between digital samples.
These are commonly called inter-sample peaks.
For example, a master might measure:
| Measurement | Value |
|---|---|
| Sample Peak | −0.5 dBFS |
| True Peak | +0.4 dBTP |
Although none of the stored samples exceeds 0 dBFS, the reconstructed signal may exceed that level.
This matters because digital-to-analog conversion and lossy encoding, such as MP3 or AAC, can introduce additional peak-related distortion.
What True Peak values should I aim for?
There is no universal ceiling that suits every production.
Some practical considerations:
| True Peak | Interpretation |
|---|---|
| −3 dBTP | Generous peak headroom |
| −2 dBTP | Conservative for further encoding |
| −1 dBTP | Common safety-oriented mastering target |
| −0.5 dBTP | Closer to full scale, less margin |
| Above 0 dBTP | Potential inter-sample overload risk |
A True Peak value above 0 dBTP does not automatically mean audible distortion, but it increases the importance of checking the playback and encoding results.
What to watch for: Use True Peak metering on finished masters, especially when preparing music for distribution, streaming, or lossy audio formats.
True-Peak-Limiting can control reconstructed peaks, but depending on the material, it may also affect the sound of transients.
The goal is to find an appropriate compromise between sound quality and peak safety.
4. PSR – Peak to Short-Term Loudness Ratio
PSR measures the difference between peak level and Short-Term Loudness.
It provides an indication of the local dynamic contrast within a moving time window.
For example:
- Peak: −0.5 dBTP
- Short-Term Loudness: −9 LUFS
- PSR: approximately 8.5
A higher PSR generally indicates more distance between the peaks and the average short-term loudness.
A lower PSR indicates a denser signal with less peak-to-loudness contrast.
How to interpret PSR
| PSR | General interpretation |
|---|---|
| 12+ | High peak-to-loudness contrast |
| 10–12 | Dynamic |
| 8–10 | Punchy and relatively loud |
| 7–8 | Dense and loud |
| 5–7 | Strongly compressed or naturally dense |
| Below 5 | Very low peak-to-loudness contrast |
These are approximate listening and analysis references, not universal quality ratings.
Some instruments, such as sustained synthesizers or heavily distorted sounds, naturally have low PSR values without excessive limiting.
Which PSR value matters most?
For loud, rhythmic music, I pay particular attention to the lower PSR readings during the loudest sections, rather than the maximum value across the entire track.
A brief dip isn’t necessarily a problem. Consistently low PSR values during a dense chorus may indicate that limiting or clipping is reducing peak contrast.
However, a high PSR doesn’t automatically guarantee a punchy snare. Transient shape, frequency balance, timing, and the rest of the arrangement also matter.
What to watch for: PSR is especially useful when comparing mastering settings and evaluating the effect of compression, clipping, and limiting on rhythmic material.
5. PLR – Peak to Loudness Ratio
PLR measures the difference between the highest peak and Integrated Loudness over the measured track.
Unlike PSR, which changes over time, PLR summarizes the peak-to-loudness relationship across the complete measurement.
For example:
| Integrated LUFS | True Peak | PLR |
|---|---|---|
| −14 | −1 dBTP | 13 |
| −10 | −1 dBTP | 9 |
| −9 | −0.5 dBTP | 8.5 |
| −8 | −1 dBTP | 7 |
PLR is closely related to the combination of loudness and peak ceiling.
At −8 LUFS and −1 dBTP, the PLR is mathematically 7. It cannot simultaneously be 10 unless the loudness or peak level changes.
How should I interpret PLR?
A higher PLR generally means a larger overall difference between peaks and perceived loudness.
A lower PLR means the track is closer to its peak ceiling on average.
However, PLR is not a direct measurement of musical quality, transient shape, or arrangement dynamics.
What to watch for: Use PLR to compare the overall peak-to-loudness relationship between mastering versions, ideally with similar peak ceilings.
For loud contemporary music, values around 7–10 are common practical reference points, but the appropriate number depends strongly on the desired loudness.
6. LRA – Loudness Range (LU)
LRA stands for Loudness Range and is measured in LU, or Loudness Units.
It describes how much the perceived loudness varies over time.
Unlike PSR and PLR, which focus on the relationship between peaks and loudness, LRA is primarily concerned with the loudness differences between musical sections.
LRA uses a statistical analysis of Short-Term Loudness and excludes certain unusually quiet and loud events.
Therefore, it is not simply the difference between the quietest and loudest moment of a track.
How to interpret LRA
| LRA | General character |
|---|---|
| 0–2 LU | Very consistent loudness |
| 2–4 LU | Mild loudness variation |
| 4–6 LU | Noticeable variation |
| 6–10 LU | Pronounced loudness development |
| Above 10 LU | Very wide loudness variation |
These are descriptive ranges, not mastering targets.
For example, a consistent Boom-Bap beat might have an LRA of 1.5 LU, while a cinematic composition with quiet passages and dramatic climaxes might reach 7 LU or more.
Neither is inherently better.
A low LRA also doesn’t mean the drums lack punch.
A track can have an LRA of 1 LU while maintaining strong kick and snare transients.
What to watch for: Use LRA to understand the loudness development of an arrangement, not to judge whether the transients are powerful enough.
7. Putting It All Together
Each measurement answers a different question.
| Measurement | What does it tell me? | Main use |
|---|---|---|
| LUFS-I | How loud is the complete track? | Overall mastering loudness |
| LUFS-S | How loud is this section? | Verse/chorus balance |
| LUFS-M | How loud is this brief moment? | Short loudness changes |
| dBFS Peak | How high are the digital samples? | Gain staging and headroom |
| dBTP | How high might reconstructed peaks reach? | Playback and codec safety |
| PSR | How much local peak contrast remains? | Short-term dynamics |
| PLR | How do overall peaks relate to loudness? | Global peak-to-loudness comparison |
| LRA (LU) | How much does loudness vary over time? | Arrangement and loudness development |
One important distinction: LU is a relative loudness unit, while LUFS is an absolute loudness level relative to digital full scale. A difference of 3 LU corresponds to a 3 dB difference on the LUFS scale.
8. Example: An Energetic Hip-Hop Master
Let’s imagine a finished beat with the following measurements:
| Measurement | Value |
|---|---|
| Integrated Loudness | −9 LUFS |
| Loudest Short-Term Section | −8 LUFS |
| Maximum True Peak | −0.5 dBTP |
| PSR in the Chorus | 8–9 |
| PLR | 8.5 |
| LRA | 3 LU |
What can we conclude?
The track is relatively loud, with a reasonably consistent loudness level across its sections.
The chorus maintains some peak-to-loudness contrast, and the overall master has been brought close to full scale.
But these values don’t tell us whether the snare sounds great, whether the bass is well-balanced, or whether the limiting introduces audible distortion.
Those are listening decisions.
What if the PSR drops to 6?
I would listen carefully to the loudest sections and investigate whether the limiting is unnecessarily flattening transients.
What if the LRA is only 1 LU?
I would check whether that matches the arrangement. If the beat is intentionally consistent, there may be nothing to change.
What if True Peak exceeds 0 dBTP?
I would check the final output and test how the master behaves during lossy encoding before making a decision.
The measurements identify potential questions. They don’t automatically provide the answers.
9. Different Music, Different Priorities
I don’t use exactly the same mastering targets for every piece of music.
My current approach distinguishes between different uses.
| Music version | Loudness approach | Peak approach |
|---|---|---|
| Artist Loud Master | Typically −8 to −10 LUFS | Optimized by ear, with codec checks |
| Artist Vocal Ready | Around −17 LUFS | Typically −3 to −6 dBFS Sample Peak |
| Artist Stems | No fixed loudness | No master-bus processing |
| Creator Music | Typically −16 to −18 LUFS | True Peak generally ≤ −1 dBTP |
| Listener Music | Same approach as Artist Loud | Music-first mastering |
These are my own production guidelines, not industry requirements.
Loud Masters
For my Loud Masters, I focus on impact, clarity, tonal balance, and the musical character of the track.
I pay particular attention to Integrated LUFS, PSR, True Peak, and the sound of the finished master after encoding.
Vocal Ready
For Vocal Ready instrumentals, the sound is already polished and tonally finalized, but the final loudness-maximizing stages are omitted.
Here, sample-peak headroom and preserved dynamics are more important than pushing the signal toward full scale.
Creator Music
For music intended for films, videos, podcasts, and other media, I prioritize controlled peaks, appropriate dynamics, and flexibility for integration with dialogue or other sound elements.
Stems
Stems are delivered without master-bus processing, allowing further mixing and mastering without inheriting the processing of the finished stereo master.
10. My Practical Mastering Checklist
When finishing a master, these are the questions I ask myself:
- Integrated LUFS: Does the overall loudness suit the music?
- Short-Term LUFS: Do the different song sections have the intended loudness relationship?
- PSR: Have I preserved enough local peak contrast, particularly in the most energetic sections?
- PLR: Does the overall peak-to-loudness relationship make sense for the chosen loudness?
- LRA: Does the loudness development reflect the arrangement?
- True Peak: Are the peaks appropriate for the intended delivery format?
- Codec testing: Does the music still sound good after MP3, AAC, or other relevant encoding?
I use meters such as Youlean Loudness Meter, ADPTR Streamliner, and the built-in metering of my mastering tools to monitor these aspects.
But the final question is always the same:
Does the music actually sound better?
Final Thoughts: The Numbers Serve the Music
Audio meters are powerful tools, but they measure specific properties of a signal, not the artistic quality of a recording.
A higher PSR doesn’t automatically mean better punch.
A higher LRA doesn’t automatically mean a more exciting arrangement.
And a louder master doesn’t automatically create a more powerful listening experience.
The purpose of metering is to understand what is happening to the sound, recognize potential problems, and make more informed decisions.
I use technical measurements to support what I hear, not to replace it.
The numbers serve the music — not the other way around.