TestBenchmark logoTestBenchmark

Screen Tearing Test

This screen tearing test slides a tall, high-contrast bar across your screen and asks your eyes one question: does the bar stay in one piece, or does it break along a horizontal line? Pick a speed, pick a frame rate that fights your refresh rate, and watch. The frame-timing panel beside it measures how evenly your browser is actually delivering frames. No download, no sign-up, nothing uploaded.

Read this first — what this test can and cannot show

A clean result here does not certify that your setup never tears. Browsers and your operating system's own compositor generally enforce their own sync on whatever is inside a browser window, which hides real tearing — a game running with VSync off can tear badly on the same screen where this page looks perfectly clean, and no web page can show that. There is no API that reports a torn frame, and this page cannot read your panel, your graphics driver's VSync setting, or whether G-Sync or FreeSync is actually active. What the test can do is show you what tearing looks like, create the conditions that make it visible, and measure how evenly frames are delivered — and the report below says exactly what was and was not shown, in those words. If you see a break, it is real. If you see nothing, it means no tear reached this window, at these settings, in this browser.

The patternNot started yet

Nothing moves until you press Start. Watch the whole bar, top to bottom: a tear cuts it at one height and shifts the lower part sideways. Faster speeds make a tear easier to see, because the step gets wider. Press F for fullscreen first — it matters more here than on any other display test, because a windowed tab is composited with sync applied and fullscreen frequently exposes what the window hides.

Space pauses · arrow keys change speed · F goes fullscreen

Speed

How fast the bar crosses the screen. At 60 Hz, 1,920 pixels a second means 32 pixels per frame — so a tear shifts the lower part 32 pixels sideways, which is easy to see. Start at 960, then try 1,920.

960 px/s · 16 px per frame at 60 Hz

Pattern

A tear is a sideways step at a horizontal boundary, so a thin line on a dark field shows it best — but a bright room or a screen where a white line is hard to follow may suit the reversed pattern. The thin line makes the same step stand out more, though the line itself is harder to follow at the top speeds.

Simulated content frame rate

Tearing appears when content updates at a rate that does not line up with the display refresh. This control quantises the bottom lane's animation to the frame rate you pick, reproducing that mismatch. A rate that divides evenly into your refresh (60 fps on 120 Hz) tends to look clean; one that does not (100 fps on 144 Hz, or 75 fps on 60 Hz) is where problems surface. The top lane always moves with the browser's own frame clock, so the two lanes show you synchronised and unsynchronised motion side by side.

Top lane — synchronised (moves with the browser's frame clock)

Bottom lane — unsynchronised (moves at the simulated content frame rate above)

The thin lines across each lane are fixed reference lines: a tear shows up as a sideways step in the bar exactly where it crosses one of them. If the bar simply advances unevenly without splitting, that is stutter, not tearing — see the comparison section below.

Frame timing — measured, not judged

While the pattern runs, the page records the interval between every animation frame the browser hands it. These numbers describe frame delivery — how evenly frames arrive — which is a different question from whether a tear is visible. Read them together with what your eyes saw: even delivery plus a visible tear points at synchronisation; uneven delivery plus a tear-like wobble usually turns out to be judder. The panel appears once 60 frames have been recorded (about a second); let it run 20 to 30 seconds for a settled reading, because short samples are dominated by warm-up noise.

No timing yet — press Start and let the pattern run.

What did you see?

The tear is something only your eyes can confirm — the page records what you report, nothing more. Pick the one that matches, and the box below says what it means. An artefact that changes with speed, with fullscreen, or with the simulated frame rate is the diagnostic one: that is tearing.

Observation log:

No observations recorded yet — run the pattern, then pick what you saw above.

Your tearing report

The report puts your observations and the measured frame timing together and says what to try next — VSync, adaptive sync, or a frame-rate cap, in that order. It never declares a setup clear of tearing: a clean result is reported as "no tear seen in this browser, in this window, at these settings", because that is all a browser page can honestly claim. Copy the text version to paste into a support ticket or a forum post when you ask for help with a game that tears.

How to run the test in 4 steps

1 · Go fullscreen first. Press “Go fullscreen” (or F). This matters more here than on any other display test: browsers often composite windowed content with vertical sync applied, which hides tearing entirely, and fullscreen frequently exposes it. A tear above or below a small window goes unseen, so the pattern needs the whole screen.

2 · Start at a moderate speed. Press Start with the default 960 pixels a second and watch the whole bar, top to bottom, for a few crossings. Here is a worked example. At 60 Hz and 1,920 pixels a second, the bar moves 32 pixels per frame (1,920 divided by 60). If two frames share the screen, the part of the line below the tear sits one frame of travel — 32 pixels — sideways from the part above it. At 960 pixels a second the step is 16 pixels, and at 2,880 it is 48. That is why a faster speed makes a tear easier to see: the step gets wider. On a 144 Hz display the same 1,920 pixels a second is only 13.3 pixels per frame, so the step is smaller and you may need the fastest speed.

3 · Watch for a horizontal split. A tear looks like the bar suddenly appearing offset — part of it shifted out of line with the rest — rather than moving as one clean shape. The fixed reference lines across the lane help: a tear shows up as a sideways step exactly where the bar crosses one of them. Let the pattern run for 20 to 30 seconds at each setting; short samples are dominated by warm-up noise, and a tear that drifts up and down the screen is easy to miss in five seconds.

4 · Change one thing and compare. Try a simulated frame rate that does not divide into your refresh rate (75 fps on a 60 Hz display, or 100 fps on a 144 Hz one), then try the same speed in fullscreen and in a window, and record what you saw each time with the buttons above. An artefact that changes with speed, with fullscreen, or with the frame rate is tearing. If nothing changes anywhere, record that too — "no tear seen" is a result, just not a certificate.

What screen tearing is, and what it looks like

Screen tearing is a horizontal break in a moving image. The screen shows the top part of one frame and the bottom part of another at the same time, so anything with a vertical edge — a doorway, a lamp post, the edge of the moving bar in this test — looks sliced and offset along one or more horizontal lines. It happens because your graphics card sends a new frame to the monitor before the monitor has finished drawing the previous one: the display reads the framebuffer top to bottom on each refresh, and if the framebuffer changes mid-read, the boundary between the old frame and the new one is the tear. Its signature is specific: the break is always horizontal, running left to right across the screen; it appears only during motion, and freezing the image makes it vanish; the seam usually drifts vertically, creeping up or down the screen, because the mismatch between frame rate and refresh rate is not a whole number; and there can be more than one seam at once when the frame rate is well above the refresh rate. If what you see does not match that description, it probably is not tearing.

Tearing, stutter, ghosting, motion blur and flicker — telling them apart

These five get confused constantly, and they have different causes and different fixes. Tearing is a horizontal break: the image is offset above and below a line, because a new frame is shown part-way through a refresh. Stutter is motion that hesitates, jumps or briefly freezes while the image stays whole, because frames arrive late or unevenly — try the FPS Test next. Ghosting is a faint trail that follows moving objects, because the panel's pixels take too long to finish changing colour — try the Ghosting Test. Motion blur is everything smearing during motion, evenly, because the image is held on screen for the whole refresh. Flicker is the whole panel pulsing in brightness, because of backlight modulation or an unstable refresh. The quickest discriminator is to pause the animation: tearing and stutter disappear completely when nothing moves, ghosting and blur smear rather than break, and flicker is the only one still visible on a static image. Choosing the right phenomenon prevents unnecessary conclusions about your hardware based on a single synthetic animation.

What causes screen tearing?

A display refreshes on a fixed schedule — 60 times per second on a 60 Hz panel, 144 times on a 144 Hz panel — and reads the framebuffer top to bottom on each cycle. Tearing happens when the framebuffer changes during that read. The specific triggers: a frame rate above the refresh rate — 200 fps on a 60 Hz panel means new frames arrive mid-refresh constantly, and this classic case can produce several seams at once; a frame rate below the refresh rate that is still unsynchronised — less obvious but just as real, because an unsynchronised 45 fps on a 60 Hz display still lands mid-refresh; VSync disabled in a game, a driver profile or a compositor setting, which is the most common root cause on the desktop; adaptive sync switched off, unsupported, or running outside its range — variable refresh fixes tearing only inside its supported range, commonly 48 to 144 Hz, and below the floor the behaviour depends on whether low-framerate compensation is available; a mismatched configuration, such as adaptive sync enabled on the monitor but not in the driver, or enabled for fullscreen but not for borderless windowed mode; and a cable or port that cannot carry the chosen resolution, refresh rate and colour depth, which may quietly fall back to a lower refresh rate. On Linux especially, tearing often comes from the desktop compositor rather than the application.

How to fix screen tearing, in order

Work through these in order. The first three solve the large majority of cases. 1 · Confirm the display is running at its intended refresh rate. Windows: Settings, then System, then Display, then Advanced display. macOS: System Settings, then Displays. Then verify with the Refresh Rate Test rather than trusting the setting alone — a very common discovery is a 144 Hz or 165 Hz monitor still set to 60 Hz because Windows, macOS or the driver defaulted it there after a cable change, a driver update or a docking-station reconnect. 2 · Enable adaptive sync everywhere it exists. It usually has to be enabled in two places: the monitor's own on-screen menu (often called FreeSync, Adaptive-Sync, G-Sync or VRR) and the graphics control panel. Enabling one without the other is the single most common misconfiguration. 3 · Turn VSync on where adaptive sync is not available. On a fixed-refresh display, VSync removes tearing outright. The cost is latency and a harsher frame-rate drop when the content cannot sustain the refresh rate. 4 · Cap the frame rate just below the refresh rate. With adaptive sync active, a cap a few frames below the maximum (for example 141 on a 144 Hz panel) keeps the content inside the variable-refresh range instead of hitting the ceiling, where VSync or tearing takes over. 5 · Match the display mode to the sync method. Exclusive fullscreen historically gave the most direct control; modern Windows borderless windowed usually works well too, but some titles behave differently in each mode — worth testing both. 6 · Check the cable and port. High refresh at high resolution needs the bandwidth to match: a DisplayPort 1.4 or HDMI 2.1 cable is required for many 4K high-refresh modes, and an older cable may silently negotiate a lower rate. 7 · Update the graphics driver, then retest at the same settings, so you are comparing like with like.

Why fullscreen changes the result

A browser window and a game reach the screen by different routes. The browser draws each frame just before the screen repaints, and most desktops show windows through a compositor that keeps to the refresh — Windows DWM, for example, composites every window with its own sync. A game running with vertical sync off bypasses that path and presents frames as fast as they are made, so it can tear where the browser cannot. Going fullscreen in this test gets you closer to the game's route: the pattern fills the screen, the compositor has less to blend, and a tear at almost any height is in view. That is also why a clean result in a window is weak evidence: test in fullscreen, at a frame rate that fights your refresh rate, before concluding the setup never tears — vertical sync may simply be doing its job.

Testing a phone or tablet display

This test runs on phones and tablets, and the controls are built for touch — but mobile deserves a word of caution. Mobile operating systems composite aggressively and manage refresh rate dynamically: many phones drop from 120 Hz to 60 Hz or lower when content is static, when the battery is low, or when the device is warm, and a browser page is rarely given an unsynchronised presentation path at all. What that means in practice: on a phone, the timing half of this test is genuinely useful — it will show you whether your device is currently running at 60, 90 or 120 Hz — while the tearing half is much less likely to reveal anything, because the compositor prevents it. For the best mobile result, rotate to landscape, turn the screen brightness up, disable any battery-saver mode, and keep the device plugged in if you are running a long measurement.

Frequently asked questions

Can a browser really test for screen tearing?

Only half of it, and this page says so up front. A browser page can create the conditions that make a tear visible — a fast-moving, high-contrast bar, at a frame rate that does not line up with your refresh rate — and it can measure how evenly frames are delivered. What it cannot do is observe your panel: there is no API that reports a torn frame, and no web page can read your graphics driver's VSync or FreeSync setting. So this test presents the moving pattern for you to judge with your own eyes and the timing numbers separately, and it never prints a pass or fail. A clean result here is genuinely good news about your desktop, and only weak evidence about your games.

Why don't I see any tearing in this test even with V-Sync off?

Because the browser window is not the game. Browsers and your operating system's own compositor — Windows DWM, for example — generally enforce their own sync on whatever is inside a browser window, which actively suppresses real tearing there regardless of your system's V-Sync setting. A game running in fullscreen exclusive mode bypasses the compositor and can tear on the same screen where this page looks clean. That is expected, not a fault in the test: try the fullscreen button, which gets you closer to how tearing behaves outside a windowed tab, and then check the setting in the game itself or in your graphics driver's control panel.

What does screen tearing look like?

A straight horizontal break in a moving image, with the picture offset above and below the line — the top part of the screen shows one frame and the bottom part shows the next, at the same time. On this page the moving bar breaks at some height, and the lower part sits sideways from the upper part. Three things mark it out: the break is always horizontal, it appears only during motion (freeze the image and it vanishes), and the seam usually drifts up or down the screen as the frame rate and refresh rate drift apart. If what you see does not match that — a trail behind the bar, a brief freeze, an overall smear — it is probably not tearing; the comparison section below tells the three apart.

What causes screen tearing?

A timing mismatch between your graphics card and your monitor. A display refreshes on a fixed schedule — 60 times a second on a 60 Hz panel, 144 times on a 144 Hz panel — reading the framebuffer top to bottom on each cycle. Tearing happens when the framebuffer changes during that read: the upper rows come from the old frame, the lower rows from the new one, and the boundary is the tear. The usual triggers are a frame rate above the refresh rate (200 fps on a 60 Hz panel means new frames arrive mid-refresh constantly), a frame rate below the refresh rate that is still unsynchronised, VSync disabled in a game or driver profile, adaptive sync switched off or running outside its supported range, or a cable that quietly negotiated a lower refresh rate than the monitor supports.

Is screen tearing harmful to my monitor?

No. Tearing is a timing artifact in how frames are presented, not a physical stress on the panel. It does not damage hardware and it does not shorten the display's life. It is annoying, especially in fast camera pans in games, and it is a sign that your frame delivery and your refresh rate are not synchronised — but it is not a fault in the monitor itself, and it is not a reason to return a panel. Fix the sync settings first; the fix guide below works through them in order.

Should I use G-Sync or FreeSync instead of V-Sync?

If your monitor supports one and it is enabled, yes, generally. VSync makes the graphics card wait for the monitor's refresh before showing a new frame, which removes tearing completely at the cost of added input lag — and if the frame rate drops below the refresh rate, VSync can halve it (on a 60 Hz display, dipping below 60 fps can drop you straight to 30). G-Sync and FreeSync turn the relationship around: the monitor's refresh rate adjusts in real time to match whatever the graphics card outputs, which removes tearing with only a fraction of VSync's input lag. Both have to be enabled in two places — the monitor's own on-screen menu and the graphics driver — and both only work inside the monitor's supported range, commonly 48 to 144 Hz. A practical tip if you use adaptive sync: cap your frame rate a few frames below your monitor's maximum refresh rate, for example 141 fps on a 144 Hz display, so the card stays comfortably inside the range where adaptive sync works best.

Why does the refresh rate show 143.8 Hz instead of 144 Hz?

Because it is measured, not read from the panel. The value is 1000 divided by the median interval between the animation frames the browser hands this page, and that interval carries a fraction of a millisecond of scheduling noise. Readings within roughly 1.5% of the nominal rate are normal and healthy. Many displays also run at a rate that is not exactly the round number on the box — 143.98 Hz, 59.94 Hz and similar are common by design. What matters is whether the reading is close to the rate you set: a 60 Hz reading on a 144 Hz monitor means something is limiting frame delivery, and the first thing to check is the operating system's display settings.

The test says 60 Hz but my monitor is 144 Hz — what is wrong?

Usually the display is genuinely running at 60 Hz. Check the operating system's display settings first — Windows, macOS and graphics drivers all reset the refresh rate after certain driver updates, cable changes and docking-station reconnections. Other causes: a cable or port without the bandwidth for the mode you selected, a laptop on battery-saver, a browser window sitting on a different (60 Hz) monitor in a multi-display setup, or a browser that limits animation frames for background or occluded windows. Move the window fully onto the fast display, plug the laptop in, and run the Refresh Rate Test again. If it still reads 60 Hz, the setting — not the monitor — is the problem.

How this test works: the page slides a tall, high-contrast bar across two lanes with requestAnimationFrame — the top lane moves with the browser's frame clock and the bottom lane moves at the simulated content frame rate you pick — and records the interval between every animation frame the browser hands it, which produces the estimated refresh rate, median frame time, stability percentage, long-frame count and 99th percentile shown beside the test. A tear is judged by your eyes only: there is no API that reports a torn frame, and the page prints no pass or fail. No downloads, no sign-up and nothing uploaded.