How accurate is a vibration meter app on your phone?

A phone vibration meter is accurate enough to answer everyday questions: is it shaking, how fast, and did my fix help? It’s good at frequency, fair at comparing readings, and rough at absolute strength. That applies to ours too: a good guide, not a certified measurement. Before you rely on your phone, run the accelerometer test. It shows what your sensor reports and how fast it reports it.

What’s inside your phone

Your phone has a tiny motion sensor called a MEMS accelerometer. It’s a speck of silicon with a small mass held on springs. When the phone moves, the mass lags behind a little, and the chip turns that into a reading. It measures along three directions: side to side, top to bottom, and in and out of the screen.

It was built to rotate your screen, count steps and run games, not to be a lab instrument. It also feels gravity as a steady 9.8 m/s² (1 g). The vibration tools remove gravity, so they only see the shaking. In a browser, the page gets about 60 readings per second, each with a timestamp.

Vibration meter app accuracy: what phones do well

Frequency

Frequency depends on timing, not on how well the sensor is calibrated. The timestamps are precise, so a phone measures how fast something shakes surprisingly well. Our engine tests feed in test signals and require a 5 Hz shake to come out within 0.1 Hz, and a 12.5 Hz shake within 0.15 Hz. They still pass when the timestamps are jittered by ±4 ms. The how it works page shows the full test table.

These tests check the math, not your phone. In real use you get a good frequency when the shake is steady, slower than about 30 Hz, and clearly stronger than the sensor’s own noise.

Comparisons

Use the same phone, in the same spot, set up the same way, and a change in the reading is real. That makes a phone very useful for:

  • Before and after a fix, like leveling a washer or balancing a tire.
  • Balancing a fan, one blade at a time.
  • A monthly check on a pump or motor, to spot a rising trend.

Catching and timing events

For a log of when shaking happens, how long it lasts and how often, a phone does the job well. None of that needs lab accuracy.

Where phones fall short

The 30 Hz ceiling

With about 60 readings per second, the fastest vibration a phone can measure is about 30 Hz (1,800 RPM). Faster shaking can show up as a false, slower frequency. This is called aliasing, and the tool warns you when it suspects it.

This matters more than it sounds. Many machine faults, like worn bearings and gears, shake far faster than 30 Hz. The machine zones in ISO 10816-1 are based on vibration from 10 to 1,000 Hz. A phone only sees the slow end of that range, so on a fast machine its mm/s reading can come out too low.

Calibration

Professional sensors are checked against a reference and come with a certificate. Phone sensors aren’t. Their scale can differ a little from phone to phone, and even between the three directions on one phone. Two phones on the same machine may disagree on strength.

Sensor noise

Every sensor jitters a little, even on a still table. The tools measure this during the 2-second “Hold still…” step and subtract it. The car, motorcycle and ceiling fan tests skip that step, because the engine or fan is already running when the measurement starts (on the fan pages you switch the fan on during the countdown). They use a fixed threshold instead: below about 0.02 m/s² (and, on the fan pages, below about 1.2 mm/s) you see “No vibration detected.” On the other tools, a reading below 1.5 times the noise gets that message. Very small vibrations, like distant traffic under a solid floor, can hide in the noise.

Chrome and Samsung Internet on Android also round each reading to 0.1 m/s² (about 0.01 g), so very gentle shaking can read as zero or jump in steps.

How the phone sits

Professionals bolt, glue or magnet-mount their sensors. A phone just rests on the surface. It can rock on its camera bump, slide on a smooth lid, or bounce on a strong shake. On a thin panel, like a washer lid, you partly measure the panel, which flexes more than the machine body.

Peak readings

The building tools show an estimated peak velocity: the RMS value × √2. That’s exact for a smooth, steady shake. For sharp single jolts, like a pile hammer blow or a door slam, the true peak is higher. Construction seismographs record the true peak in each direction.

Very slow motion

Slow sway, like a tall building in the wind or a dryer drum turning under once a second, has tiny acceleration. It sits close to the noise, so the velocity is hard to pin down, and most tools ignore anything below 1 Hz. The ceiling fan tests look down to 0.5 Hz, because a fan’s slow wobble is still strong.

The browser

The browser can deliver fewer readings when the phone is busy, hot or in battery saver. It stops sending them when the screen locks or the tab goes to the background. The tools measure the real rate every time and show it in the details, so the frequency scale stays right. But a slower rate also lowers the limit, and shaking faster than the new limit can show up at a false, lower frequency and read too strong. Under about 40 readings a second, trust the reading only for slow movement. Firefox on Android sends only about 10 readings a second, so use Chrome or Samsung Internet there.

Phone vs professional vibration meter

Phone in a browserProfessional meter or seismograph
Frequency rangeAbout 1 to 30 HzMuch wider, often hundreds of Hz or more
CalibrationNoneChecked against a reference, with a certificate
MountingRests on the surfaceBolted, glued, magnet-mounted or fixed to the ground
Peak valueEstimated from RMSTrue peak in each direction
Accepted in a disputeNoYes, with a proper report
CostFree, already in your pocketA real purchase, or a paid service

How to check your own phone

These checks take ten minutes and tell you how far to trust your phone.

  1. Gravity check. Open the accelerometer test and lay the phone flat and still. One axis should read about 9.8 m/s². Then stand the phone on its long edge, then its short edge, against something solid. Each time, the axis pointing down should read about 9.8. If one axis reads clearly lower, say 9.2 while the others read 9.8, readings on that axis run about 6% low.
  2. Sample rate. Note the rate and the highest measurable frequency. About 60 readings per second and 30 Hz is normal.
  3. Quiet check. Lay the phone on a solid table in a quiet room and run the vibration frequency analyzer. It should say “No vibration detected” or show only very small peaks.
  4. Known frequency check. Many washers show their spin speed. A 1,200 RPM spin is 20 Hz. Put the phone on top during the final spin: the main peak should land close to 20 Hz. The drum’s real speed can differ a little from the setting. The pendulum experiment is another good test: if your value of g comes out close to 9.81 m/s², the timing works.
  5. Repeat check. Measure the same steady machine three times without moving the phone. The readings should be close. A big spread means the phone is shifting, or the machine keeps changing speed.
  6. Two-phone check. If you have a second phone, lay both side by side on the same surface. The difference between them shows how much phones vary.

How to get the most reliable reading

  • Lay the phone flat, screen up, on a hard, flat spot. Avoid thin lids, rubber mats and cushions.
  • Take off a thick or loose case if the phone rocks. On a smooth or sloped machine, tape the phone down.
  • Keep your hands off, and keep still during the “Hold still…” step.
  • Let the test run its full length: 30 seconds on most tools, 60 on the analyzer.
  • Keep the screen on and close other tabs. Plug in for long logs.
  • When comparing, use the same phone, the same spot and the same settings.
  • Take the aliasing warning seriously. If it appears, the real vibration is probably faster than the phone can measure.
  • Tap “Save image” or “Download CSV” each time, so you have a record.

When a phone isn’t enough

  • Disputes and claims. For construction, landlord or insurance disputes, hire a professional vibration monitoring company.
  • Fast machines. For bearings, gearboxes, and motors above about 1,800 RPM, a technician with a proper vibration analyzer is the right call.
  • Structural worries. For cracks, sagging floors or anything that looks unsafe, ask a structural engineer.
  • Safety-critical equipment. Never rely on a phone where a wrong reading could hurt someone.

Used within its limits, your phone gives quick, honest answers. Check your sensor first with the accelerometer test, then pick the tool for the thing that shakes.

Test it yourself

Put your phone flat on the thing that shakes and tap Start. You get a reading, the frequency and a plain-English label in about 30 seconds. Free, no app, and your readings stay on your phone.

Start the test: Accelerometer test

Frequently asked questions

Are vibration meter apps accurate?

For frequency, often yes: a steady shake below about 30 Hz usually comes out within a fraction of a hertz. For strength, they're good for comparing readings on the same phone, but two phones can disagree. Treat any phone reading as a good guide, not a certified measurement.

Is a vibration meter app more accurate than a website?

A native app can often read the sensor faster than a browser allows, which raises the highest frequency it can measure. The sensor, the lack of calibration and the way the phone sits are the same, so strength readings have the same limits. A website needs no install, and this one keeps your readings on your phone.

Why do two phones give different readings on the same machine?

Their sensors differ slightly in scale and noise, their cases differ, and they sit on slightly different spots. Frequency should agree closely. Strength can differ more, so use one phone for all the readings you want to compare.

Does my phone case affect the reading?

It can. A thick, soft case can soak up some of the shake, and a loose case can rattle and add false peaks. A slim, snug case is usually fine. If in doubt, measure once with the case and once without.

Can I use a phone reading in a dispute?

As a dated record of when shaking happened and roughly how strong it was, it can help you start a conversation. It isn't proof. For insurance claims or legal disputes, hire a professional who uses calibrated instruments.