For about a year I had a screen that gave me a headache in ten minutes flat. My phone. Not every phone — this one, and only sometimes. I'd doom-scroll in bed at night and come away with a dull ache behind my eyes and blurry text. During the day, on the same phone, nothing. I assumed I was just tired, or getting old, or staring too long.
I was wrong on all three. The culprit was PWM flicker — a dimming method that switches your backlight on and off hundreds or thousands of times a second. About 70% of people can't consciously see it, which is exactly why those of us who feel it spend years quietly doubting ourselves.
I wrote the full, source-backed version on my site, with the embedded test tool, the annotated camera figure, and the IEEE-1789 safety numbers:
👉 PWM Flicker Test: Is Your Screen Causing Eye Strain?
This DEV.to version keeps the practical parts — how to actually test for it and what to do about it.
What PWM flicker actually is
When you turn screen brightness down, the panel has to emit less light. There are two ways to do that:
- DC dimming lowers the actual current, so the light stays continuously on at a lower level. No flicker.
- PWM keeps the light at full power but switches it on and off very fast. Your average brightness drops because the light is dark part of every cycle.
That on/off strobing is invisible to most people. But for the ~10–20% of us who are PWM-sensitive, the brain still processes the flicker, and the result is headaches, eye strain, dizziness, and blurry text after a screen session.
The maddening part is the brightness correlation. PWM is strongest at low brightness and usually disappears at 100%. So the one phone that wrecks your eyes at night is fine in daylight when you crank it up — which makes the whole thing feel like it's in your head.
If you've used phones happily for a decade and only struggle now, your eyes didn't change — the dimming technology did. OLED-everywhere in 2025–26 is the reason complaints are exploding: OLED pixels don't dim cleanly by lowering current, so manufacturers reach for PWM.
The 5-second self-check (no tools)
Before anything else, do the stroboscope test. Hold a pencil — or just spread your fingers — a few inches in front of the screen and wave them quickly side to side.
- Smooth blur → likely flicker-free at this brightness.
- Several distinct, ghosted copies → that's PWM chopping your hand into snapshots.
It's crude, but it's a real, instant signal. If you see ghost copies, keep reading.
Testing it properly in the browser
The tool I built has two modes, and they solve two different problems.
It auto-detects your display's refresh rate when it loads, which matters because of one honest limitation I want to be upfront about:
A web page can only flicker up to half your display's refresh rate. So on a 60 Hz screen, the on-screen flicker maxes out around 30 Hz — that's a reference pattern, not your monitor's real PWM frequency.
I could have hidden that and pretended the page measures your panel directly. It doesn't, and neither can any web page. That's the whole reason the camera method exists.
Mode 1 — FLICKER (film it with a phone)
This mode flashes the screen, honestly capped at about half your refresh rate, and the workflow is: go fullscreen, then film the screen with a second phone. Your monitor's real PWM frequency shows up as rolling dark bands in that footage. It's a calibration aid — it teaches your eye and your camera what banding looks like before you point the camera at the device you actually suspect.
Mode 2 — MOVING LINE (no camera at all)
This is the one I reach for first. A bright bar sweeps across the screen at your display's native refresh rate. If your panel is clean, you see one smooth bar. Under PWM, your eyes break that bar into several separated copies instead of one smooth blur — because the light is only on for slices of each sweep.
No second device, no exposure fiddling. If you see the bar split into copies, you're sensitive and your screen is flickering.
The phone-camera method (for measuring your actual monitor)
To find your monitor's real PWM frequency — not the web reference — you film it. Here's the part that took me longest to get right:
- Drop the test screen to low brightness — around 20–40% for phones, 50–75% for monitors. PWM is strongest here. At 100% most screens go steady and you'll see nothing.
- Lock exposure on the camera. iPhone: tap-and-hold to get AE/AF Lock. Android: use Pro/Manual mode.
- Use a fast shutter — somewhere from 1/120s up to 1/4000s. A fast shutter is what freezes the flicker into visible bands.
- Read the bands:
Thick, slow, dark bands = low frequency = worse.
Thin, faint, fast lines = high frequency = safer.
No bands at all = DC dimming or a true flicker-free panel.
How bad is "bad"? The frequency safety scale
Count the bands and you can ballpark where your screen lands. Based on IEEE 1789-2015 flicker guidance:
240–480 Hz → Poor (high eye-strain risk)
960–1440 Hz → Fair (still a higher-risk zone)
1920–2880 Hz → Good (comfortable for most sensitive people)
3840 Hz+ or DC dimming → Flicker-free (no meaningful risk)
One nuance most guides skip: modulation depth matters alongside frequency. A screen that goes fully dark each cycle (100% depth) is harsher than one that only dips part-way at the same frequency.
What to actually do about it
If you've confirmed PWM is the problem, here's the fix order from free to spendy:
- Raise brightness above the PWM threshold. Often the fastest relief — many screens go steady near full brightness. (Pair it with a darker room or night mode so it's not blinding.)
- Enable DC dimming / anti-flicker if your device has the toggle (a lot of newer Androids do, sometimes called "anti-flicker" or "DC dimming" in display settings).
- Take real breaks — the 20-20-20 rule: every 20 minutes, look 20 feet away for 20 seconds.
- Buy with flicker in mind. High-frequency-PWM phones (Honor's Magic line runs ~3840–4320 Hz; OnePlus 15 ~2160 Hz) or TÜV Rheinland Flicker-Free certified monitors. Quality IPS LCD panels with DC dimming are still the safest bet if you're badly affected.
The single most useful thing this whole rabbit hole gave me wasn't a fix — it was confirmation. Seeing the bright bar visibly split into copies on my "headache phone," and a smooth bar on my work monitor, was the moment I stopped second-guessing myself. If a screen makes your head hurt and others shrug, you're probably not imagining it.
Quick FAQ
Why does only one of my devices give me a headache?
Different panels use different PWM frequencies and depths. A low-frequency, 100%-depth OLED can wreck a sensitive person while a DC-dimmed LCD next to it feels fine.
Why is it worse at low brightness?
PWM dims by spending more of each cycle dark. The lower the brightness, the longer the "off" portion — so the flicker is most aggressive dim and often vanishes at 100%.
Can a website measure my monitor's real PWM frequency?
No — a web page can only flicker up to half your refresh rate. The on-screen test is a reference; to measure your panel's true frequency you film it with a phone and count the bands.
I don't have a second phone — can I still test?
Yes. Use MOVING LINE mode. If the sweeping bar splits into several separated copies instead of one smooth blur, your screen is flickering — no camera needed.
Is PWM dangerous or just annoying?
For most people it's harmless and invisible. For the sensitive 10–20% it causes real symptoms — headaches, eye strain, blurry text. It's not damaging your eyes long-term, but it's not "in your head" either.
Full version — with the embedded test tool, the IEEE-1789 breakdown, and the annotated camera figures — is on my site:
👉 PWM Flicker Test: Is Your Screen Causing Eye Strain?
Want to just check your screen right now? The PWM Flicker Test is free and runs in your browser — start with MOVING LINE mode at low brightness.
Top comments (0)