How is the average level calculated here?

The square root of the mean of the squared sample values, which is a way of expressing average energy rather than momentary amplitude. Because the squares are averaged, a long quiet passage and a short loud one contribute in proportion to their duration and their energy, which is closer to how loudness is perceived.

What is a typical crest factor?

A pure sine wave sits at three decibels, which is the theoretical minimum for a single tone. Heavily limited masters often land between six and ten decibels, natural speech and percussive material frequently exceed twelve, and sparse recordings with wide dynamics can go higher still.

Why is the bar chart scaled to its own maximum?

So that the shape of the level over time stays visible whatever the file's absolute level. The cards carry the absolute numbers, and the chart answers a different question: does the recording stay at one level, or does it move?

Will this tell me if my file is suitable for a platform?

Not directly, because platform normalisation is based on perceptual loudness rather than RMS. Use this meter to understand the file's level and peaks, and the loudness meter to compare against the targets services actually apply.

Why does the crest factor matter at a limiter?

Because the limiter has to reduce peaks, and a high crest factor means the peaks are far above the average. That forces more gain reduction for the same overall level change, which is why two files at the same RMS can sound and measure very differently after limiting.

RMS Meter

You Might Also Need

Crest Factor Meter

Measure peak-to-RMS ratio

VU Meter

Classic VU meter visualization

RMS Normalizer

Normalize audio using RMS levels

Peak Limiter

Prevent audio peaks and clipping

RMS Meter reports the average level of a file, its peak, and the ratio between them, then plots that average over time in half-second windows. It is the plain level measurement that most other readings are built on.

What is RMS Meter?

Four cards summarise the file — average RMS, peak, crest factor and duration — and a bar chart underneath shows how the average level moves through the recording.

What is measured and how it is presented:

RMS in dBFS together with the underlying linear value, so the root-mean-square level can be read either way.
Peak level in dBFS with its linear value, which is the largest absolute sample in the file.
Crest factor in decibels, computed as peak divided by RMS, which quantifies how peaky the material is.
A bar for every half-second window, scaled so the largest window fills the full height of the chart, with each bar's exact level available on hover.

Root mean square is the square root of the mean of the squared samples, which is why it tracks perceived energy far better than a peak reading does: a passage that is loud throughout shows a high average, while a quiet recording with one loud click does not. Squaring is what makes the measure insensitive to sign, so phase and polarity have no effect on it. The crest factor then falls out of the comparison, and it is the number that explains why two files at identical RMS can behave completely differently at a limiter — one sitting near its average and the other spiking well above it. The windowed timeline uses half-second spans, a compromise between stability and detail: short enough to show a quiet verse or a loud chorus, long enough that individual drum hits do not dominate. Bars are scaled to the loudest window rather than to an absolute reference, so the chart shows the shape of the level over time and the cards carry the absolute figures. Every figure on the panel is a plain measurement with no frequency weighting applied, which makes it a fair comparison between two files and a poor predictor of how a streaming platform will treat them. The pairing of average and peak is the point of the display: read alone, either number can mislead, and how a limiter will behave depends on the relationship between the two rather than on either one by itself.

What the numbers do not cover:

RMS is not perceptual loudness, so it does not account for frequency weighting and will not match a platform's normalisation.
The windowed chart is scaled to its own maximum, so the same visual height can mean different absolute levels in different files.
Only the maximum window is highlighted implicitly by relative height, with no marker for the quietest one.
Nothing is measured per channel, so a file with one silent channel reports a combined figure over all channels.
No hearing-weighted meter is offered, which is what the loudness meter is for.

Use it as the first level measurement on any file, and as the explanation when a limiter behaves unexpectedly: a high crest factor tells you that the peaks are far above the average and that dynamic control will matter more than gain. When the same information is needed in a perceptual form, the loudness meter and the VU meter answer related questions with different weighting.

How to use RMS Meter

Drop an audio file onto the panel; the measurement has no options to configure.
Read the RMS and peak cards together to judge both the average and the worst sample.
Check the crest factor to see how far the peaks sit above the average level.
Scan the bar chart for sections where the average level drops or rises, hovering a bar for its exact level.

When to use RMS Meter vs related tools

RMS Meter is the general-purpose level measurement, useful whenever average and peak need to be seen together with the ratio between them. Crest Factor Meter explores that ratio window by window and highlights the most peak-heavy moment, and VU Meter presents the same average with the timing and scaling of a classic needle. For perceptual loudness, Loudness Meter (LUFS) is the meter that matches platform behaviour.

Privacy & Security

This tool runs entirely in your browser — no data ever leaves your device. There is no server round-trip, no upload, no logging, and no account required. Your input is processed locally using client-side JavaScript and is never stored, transmitted, or accessible to anyone else. When you close the tab, everything disappears.