Screen flicker test:
check your monitor for PWM

First, what a browser can and cannot test

A web page can only set pixel values. Pulse width modulation happens in the backlight driver, one layer below anything JavaScript or CSS can reach, so no website can measure your monitor's PWM frequency or modulation depth. Any page that claims to report a number in hertz from inside a browser is guessing. The browser is also stuck at your refresh rate, typically 60 to 120 Hz, while backlight PWM commonly runs from a few hundred hertz to tens of kilohertz.

What a browser can do is show you fast, high contrast patterns that make the flicker visible indirectly. A backlight that switches on and off is a strobe, and a strobe turns smooth motion into a series of frozen copies. That is a stroboscopic artifact, and it is evidence rather than measurement. Treat the two tests below as a screening step, then confirm with the phone camera method, which records the light itself.

Test 1: the moving bar test

A white bar sweeps horizontally across a black field. On a flicker free display the bar smears into one continuous blur, because the light is on the whole time your eye is tracking it. On a display using pulse width modulation the backlight is only on in pulses, so the bar is lit at a series of discrete positions and breaks into several separate copies with dark gaps between them. Set your hardware brightness low before you run it, because that is where PWM is deepest.

What to look for

• One bar, blurred. The bar looks like a soft continuous streak. That is the expected result on a flicker free backlight or one running PWM fast enough that the copies merge.
• Several distinct bars. You see two, three or more crisp copies of the bar at once, separated by black. That is a stroboscopic artifact and it points to a pulsing backlight.
• Result changes with brightness. Run it at 100 percent hardware brightness, then at your lowest comfortable setting. Copies that appear only when you dim are the classic PWM signature.
• Sweep your eyes across the screen. Moving your gaze rather than tracking the bar makes the copies easier to see, the same way waving a pencil under a fluorescent lamp does.

One caveat worth knowing: sample and hold panels, frame rate limits and monitor overdrive can all add motion artifacts of their own. A single ambiguous result is not a verdict. That is what the camera is for.

Test 2: the brightness stepper

Pulse width modulation is a dimming mechanism, so it is deepest where the dimming is deepest. This test gives you a large flat gray field to watch while you walk your hardware brightness down. Use your keyboard brightness keys or your monitor's own brightness control, not a software slider, because only the hardware control changes the backlight.

1. Set hardware brightness to maximum and look at the field below. Note how steady it looks.
2. Lower hardware brightness two steps at a time, pausing at each step for a few seconds.
3. Watch the edges of the field and your peripheral vision, where flicker sensitivity is highest. Shimmer, a fine buzz, or a sensation that the field is not quite still are the signs to note.
4. Step the on screen gray darker with the buttons and repeat. A darker gray at high backlight often feels calmer than a bright gray at low backlight, which is the whole argument for software dimming.

If the discomfort tracks your hardware brightness and disappears when you raise it, that pattern is consistent with PWM. If the discomfort tracks how bright the image is regardless of the backlight setting, you are more likely dealing with overall luminance or spectrum rather than flicker.

The reliable test: your phone's slow motion camera

This is the method to trust. A camera sensor reads its rows one after another rather than all at once, so a backlight that is switching on and off gets recorded as dark bands rolling through the frame. High frame rate video makes those bands easy to see. It is not a calibrated instrument, but it observes the actual light output instead of inferring from what your eyes report.

On iPhone

1. Set the screen you are testing to a low hardware brightness, around 20 to 30 percent, and display a plain white or light gray window.
2. Open the Camera app and swipe to the Slo-mo mode.
3. Tap the frame rate indicator and choose 240 fps if your model offers it. 120 fps still works and catches lower frequency flicker.
4. Hold the phone about 30 to 50 cm from the screen, fill the frame with the bright area, and record five seconds.
5. Play the clip back and scrub slowly. Dark horizontal bands drifting through the frame mean the backlight is pulsing. A clean, evenly lit frame means no flicker at a frequency the camera can resolve.
6. Repeat at maximum brightness. Bands that vanish at 100 percent confirm the flicker is coming from backlight dimming.

On Android

1. Set the tested screen to low hardware brightness and show a plain white window.
2. Open the camera and switch to Slow motion mode. On Samsung it is under More, on Pixel it is Slow Motion in the mode list.
3. Pick the highest rate offered, usually 240 fps, sometimes 480 or 960 fps in a short burst mode.
4. Fill the frame with the bright area and record a few seconds.
5. Scrub the playback. Rolling dark bands mean pulse width modulation. If your camera app exposes a manual shutter speed, a fast shutter such as 1/1000 s makes the banding much clearer.

If you find PWM, stop dimming the backlight

The fix follows directly from the mechanism. PWM flicker gets deeper as the backlight duty cycle drops, so the way out is to leave the backlight high, often at 100 percent, and get the darkness you want somewhere that does not pulse. Software dimming does exactly that: it works on pixel values or the display gamma table, so it lowers perceived brightness with a steady backlight and adds no flicker of its own. The honest trade off is that heavy gamma dimming compresses contrast, so very dark levels lose some separation in the shadows.

Nox is the macOS app we build for this. It dims below the system minimum at the gamma level, so you can run the backlight at full brightness and still get a screen that is comfortable at night, and it ships with 12 built-in presets for warmth, tint and grayscale alongside the dimming. It is $5 once, with a 14-day free trial, and requires macOS 14 or later. If your discomfort turns out to be spectrum rather than flicker, the same presets cover that too, and the research behind them is written up here.

What pulse width modulation actually is

An LED backlight is not easy to dim smoothly by lowering its current, because the color and efficiency shift as the current drops. So most displays dim by switching the LEDs fully on and fully off, hundreds or thousands of times per second, and varying the fraction of each cycle they stay on. That fraction is the duty cycle, and the technique is pulse width modulation. At 100 percent brightness the light is essentially always on. At 20 percent it is on for roughly a fifth of each cycle and off for the rest, which is where the modulation depth becomes large enough to matter.

Two numbers describe the result: the frequency, in hertz, and the modulation depth, which is how far the light swings between its peak and its trough. Both matter. IEEE 1789-2015, a recommended practice for modulating current in high brightness LEDs, gives the framework most people cite: it defines low risk and no observable effect regions where the permitted modulation depth rises with frequency. It is an engineering document about LED lighting, not a clinical finding, and it is explicit that individual sensitivity varies.

Why some people notice and others do not is not fully settled. Flicker above the frequency you can consciously perceive still reaches the visual system, and stroboscopic effects, phantom array artifacts and eye movement interactions have all been proposed as routes by which invisible flicker produces visible or physical symptoms. Reports of headache, eye strain and discomfort from flickering light sources are common enough to take seriously, and people with migraine often describe themselves as more sensitive. We are describing an association here, not a proven cause, and this page is not medical advice. If your symptoms are persistent or severe, that is a conversation for a clinician.

Frequently asked questions

Can a website measure monitor PWM?

No. A web page controls pixel values, not the backlight, and pulse width modulation happens in the backlight driver below anything a browser can reach. The browser also refreshes at 60 to 120 Hz, far slower than a typical PWM cycle. What the tests on this page can do is reveal stroboscopic artifacts caused by that flicker, which is evidence rather than measurement. A phone camera shooting slow motion is the closest thing to a real measurement most people have.

How do I test my screen for flicker?

Set your hardware brightness low, run the moving bar test, and look at whether the bar smears into one continuous blur or breaks into separate copies of itself. Then confirm with a phone: record the screen in slow motion at 240 frames per second and look for dark bands rolling through the frame. Agreement between the two is a reasonable result. The phone result is the one to trust.

Why does PWM flicker only appear at low brightness?

Pulse width modulation dims by switching the backlight fully on and fully off many times per second and varying how long it stays on. At full brightness the backlight is on almost continuously, so there is little to see. As you dim, the on time shrinks and the off gaps grow, which increases the modulation depth. That is why a display can look perfectly steady at 100 percent and feel harsh at 20 percent.

What flicker frequency is considered safe?

IEEE 1789-2015 is the engineering standard usually cited here. It is a recommended practice for LED lighting rather than a medical finding, and it sets risk thresholds that depend on both frequency and modulation depth: the higher the frequency, the more modulation is tolerated before flicker is considered a risk. Low frequency, high depth flicker is the worst case in that framework. Reported sensitivity varies widely between individuals, so a display inside the low risk region can still bother some people.

Do MacBook screens use PWM?

Apple does not publish backlight dimming specifications, and behavior differs by model, panel type and firmware. Independent measurements of recent MacBook LCD and mini LED displays generally report no visible low frequency flicker, while OLED panels in phones and some laptops commonly do use pulse width modulation. Rather than trust a general claim about a product line, test the unit in front of you with the phone camera method.

What can I do if my monitor uses PWM?

Raise the hardware brightness until the flicker becomes shallow, often near 100 percent, and get the darkness you want from software dimming instead. Software dimming works on the pixel values or the display gamma table, so it lowers perceived brightness without touching the backlight duty cycle and does not add flicker of its own. The trade off is slightly reduced contrast at heavy dimming levels.