What are L10 and L90 — and how do you measure them with a phone?

Recently someone asked me:

“How do I measure L90 and L10 values?”

If you’ve ever Googled that question, you probably found some dense environmental acoustics documents filled with charts and complex language.

Let’s walk through this step-by-step.

The simple explanation (no acoustics degree required)

Let’s say that you have sporadic noise issues in your neighborhood, but you’re never sure exactly when they happen or how loud they can get. So you measure the sound outside your home for 10 minutes with a decibel meter. The noise goes up and down the whole time.

Instead of asking:

  • What was the loudest moment? or
  • What was the average sound level?

When you calculate the L10 and L90, you ask a different question: “For what percentage of time was the sound above a certain level?”

  • L90 = The sound level that was exceeded 90% of the time
    This is usually your background noise level

  • L10 = The sound level that was exceeded 10% of the time
    This represents recurring, louder spikes

So:

  • L90 ≈ a steady hum (HVAC, distant traffic, refrigerator)
  • L10 ≈ louder events (cars passing, leaf blowers, aircraft)

Why would anyone calculate this percentage?

Most homeowners or renters don’t need L10 and L90. But they’re useful when:

  • Filing a noise complaint
  • Documenting traffic noise
  • Showing that background noise is steady, but spikes are intrusive or stress-inducing
  • Comparing before/after mitigation

Here’s a hypothetical example:

  • L90 = 48 dB
  • L10 = 63 dB

What this data tells you:

  • The steady background (L90) sits around 48 dB
  • Regular spikes (L10) reach 63 dB
  • There’s a 15 dB swing between the baseline and events

That swing can be meaningful. For example, a 10 dB difference is perceived roughly as “twice as loud.”

A slightly deeper explanation

The average sound level (LAeq) is useful — but it can hide important details.

Let’s say that a loud truck drives by during a 10-minute measurement. That single event raises the average. But does it represent your normal background noise?

No.

That’s why environmental noise studies often use L90:

  • It filters out short spikes.
  • It gives you the steady baseline.

And L10 captures how intrusive the random louder events are.

The difference between L10 and L90 is sometimes used to describe how “variable” or unstable a noise environment can get. So a small difference = a steady environment. And a large difference = a spiky, intrusive environment.

Can the NIOSH app measure L10 and L90?

I’ve written about the NIOSH Sound Level Meter app before. It’s excellent. But it doesn’t provide L10 or L90 values.

The NIOSH app reports:

  • LAeq
  • Max level
  • LCpeak
  • TWA
  • Dose

Those are summary statistics. You can’t calculate L10 or L90 from just average and maximum values. To calculate percentiles, you need the full distribution of measurements over time.

What you actually need

To calculate L10 and L90, you need:

  • Continuous logged measurements
  • Time-stamped data
  • A spreadsheet (Excel or Google Sheets)

I’ll show you how to do this using Decibel Meter, an iOS app developed by Ashraf Thoppukadavil.

Important: the Decibel Meter app saved data displays a graph, but it exports text

While you’re using the app and you tap the “Records” button, you’ll see a graph that displays the data over the duration of your measurement. But when you export/share that measurement from the archive, you receive a text file containing detailed, moment-by-moment readings as a numerical list.

Many people don’t realize this. Here’s a snippet of what that export looks like:

Sample noise data from the Decibel Meter app

This is what you can see when you get a report from Decibel Meter:

  • Individual sound levels
  • Time stamps
  • Weighting type (in this example: dB-C)
  • Running averages

That’s exactly what we need.

How to calculate the L10 and L90 in Excel

Let’s say you record 10 minutes of sound.

Step 1: Export the file

Email it to yourself, or AirDrop it to your computer.

Step 2: Open in Excel

Copy the column with the actual dB readings into a spreadsheet. You only need the numeric values. You need to use consistent data to calculate both the L10 and L90 values, and so — in this example using the data from the Decibel Meter app — we’re going to use the “Average” dB data for both. (Don’t use the “Peak” data for this calculation!)

Step 3: Use the Percentile function

To calculate:

L10:

=PERCENTILE.INC(A1:A1000,0.9)

L90:

=PERCENTILE.INC(A1:A1000,0.1)

Adjust the range(s) to match your data.

Soundproofist demo of how to calculate L10 and L90 noise values

In this example using Google Sheets, the Average dB was listed in Column C, and had 3,462 entries in it. Selecting the range from C2:C3462, I entered the Percentile formula to get L10 (63.11) and L90 (36.74).

These results are in decibels. In other words, during this measurement period, the sound exceeded 63.11 dB for 10% of the time (L10), and exceeded 36.74 dB for 90% of the time (L90). That’s why L90 is often called the “background” level — it represents the level the environment rarely drops below.

So, L90 = the decibel level that was exceeded 90% of the time.

One important detail: A-weighting vs C-weighting

In the example export above, the data shows dB-C values. For environmental and community noise studies, L10 and L90 are often calculated using A-weighted measurements (dBA). For example, that’s what police departments use if you make a noise-related complaint. So before you measure, make sure your app is set to the appropriate weighting.

Do most homeowners / apartment-dwellers need this?

Usually, no. For typical soundproofing issues, you mainly need:

  • LAeq (average level)
  • Peak level
  • Inside vs outside comparison

L10 and L90 are more useful when:

  • Dealing with local ordinances
  • Documenting recurring disturbances
  • Working with consultants
  • Demonstrating variability in a noise environment

But if you ever need them, now you know how to calculate this information using just a phone app and Excel.

I’ll be making a short video

I’ll put together a short, clear video to walk you through:

  • What L10 and L90 mean
  • How to export logged data
  • How to calculate percentiles in Excel or Google Sheets

And I’ll embed it here once it’s ready.

Learn more:

How to calculate noise levels over time with a phone app (L10 and L90)

Leave a Reply

Your email address will not be published. Required fields are marked *