Controller testing FAQ

A deadzone tells software to ignore part of a stick’s input, usually near the center. Small movements inside an inner deadzone produce no movement in that software. It changes the response, not the physical condition of the joystick.

There is no universal best value, including for FPS games. Start with your game’s default, check untouched input, then try small changes while testing slow deliberate aim. Zero can expose unwanted movement; a larger value can discard useful small movements. Use the smallest setting that feels stable and controllable in that specific game, not a copied percentage.

Drift is unwanted stick input while the stick is released. A deadzone is a rule that ignores some input. Increasing it may stop visible camera movement without removing the underlying offset or repairing the controller.

No. Calibration concerns how position is interpreted, such as center or range. A deadzone ignores some input. Our separate browser preview illustrates both ideas but changes neither hardware nor game settings.

An inner deadzone ignores input near the center. An outer deadzone can make full output arrive before the stick reaches its physical edge. Anti-deadzone adds a starting output after movement begins, often to overcome a downstream deadzone. Names and formulas vary between software, so check what the actual control does.

Connect the controller to the device running this site, press a button, select it and start live view. Check that the axes follow the intended sticks. Release both sticks on a stable surface, then run the resting check. Repeat the observation and compare it with the symptom in your game; the browser cannot identify a failed component.

Check both sticks because their behavior can differ. This site captures the selected left and right axis pairs in the same resting run and reports each separately. Confirm each mapping before starting, especially on a nonstandard controller.

Use a supported connection that stays stable. A known data-capable USB cable can help isolate wireless connection trouble, but it does not bypass all controller processing. Keep the connection the same for comparisons, or record when it changes. Not every controller supports Bluetooth.

A stable surface helps you avoid accidentally moving or loading a stick during the resting capture. Make sure the sticks are clear of nearby objects and release them fully. The tool cannot tell whether a large reading came from your hand or the controller.

Release them before starting the resting check and leave them untouched through its three-second countdown and ten-second capture. The countdown is not included in the reported metrics. A completed run describes that interval, not every future moment of use.

The capture needs uninterrupted, focused input. Leaving the tab, losing focus, disconnecting, changing mappings or encountering a reading gap over 250 milliseconds interrupts it. This avoids joining separate or stale intervals into one apparently complete result. Return to the page and start a new run.

Start a fresh resting check with both sticks released. The tool deliberately keeps large readings rather than guessing that they were accidental. A run that includes your movement is not a useful untouched-stick baseline.

Yes. A completed resting run offers local JSON and PNG downloads. Save the model, connection and settings alongside them so later comparisons have context. Results are not stored in a cloud account, and the exported summary is not a certified hardware diagnosis.

They are the horizontal and vertical axis values exposed by the browser for the selected stick. Standard axes normally range from −1 to 1, with values near zero at center. Signs and axis assignments can differ on nonstandard mappings; move one stick at a time to confirm them.

Radial offset describes the reported position’s distance from zero: 100 × √(X² + Y²). It is relative to a single-axis full-scale value of 1. Center bias reports average signed X and Y values, which can help describe a consistent direction. Neither is a percentage of damage.

The formula combines two axes. If both report 1, the radial offset is about 141.4%, because the diagonal is longer than either single axis. We do not clamp stored resting metrics to 100%. This does not mean the controller is more than 100% damaged.

No. There is no universal browser offset percentage that certifies a controller as healthy or faulty. Repeat the released-stick check and compare actual unwanted movement in your games. Mapping, processing, centering and game settings all matter.

A game may ignore small input through a deadzone or process it differently. The browser displays the values available to it without our preview correction, but those are not guaranteed to be untouched sensor readings. A quiet game does not prove the tester is wrong.

Games can use different deadzones, response curves, bindings and input paths. Compare the relevant settings one at a time and keep a note of the originals. The same controller input can cause visible movement in one game and none in another.

Mean offset is the time-weighted average radial distance during the capture; peak offset is the largest observed radial distance. A brief spike can raise the peak much more than the mean. Average signed X/Y bias adds directional context and is not the same calculation as mean radial offset.

No. It observes inputs supplied by the browser, not the internal parts. Sensor wear, centering issues, contamination and processing can produce similar symptoms. Saved readings can support a repair conversation, but inspection or model-specific diagnosis may still be necessary.

First identify the layer you want to change: the game, a supported controller profile or remapping software. Use that software’s current controls and change one setting at a time. Xbox hardware recalibration is a separate feature with model requirements. Our browser preview does not save settings to Windows, Xbox Accessories or a game.

A deadzone mainly affects when input begins near center. Sensitivity, response curves, acceleration and game behavior also affect how movement feels. This tester does not measure end-to-end delay, so a slow feeling alone is not evidence of controller latency.

Investigate deadzone when unwanted movement occurs near center or small movements do not start responding. Investigate sensitivity when the speed or scale of deliberate movement feels wrong. Record the original values and change one control at a time so you can tell what helped.

Yes. Ignoring the offset can stop visible movement in a particular game while the same input remains present elsewhere. It can also reduce fine control. A bigger deadzone is a response adjustment, not evidence of a hardware repair.

Not necessarily. The sticks can have different resting behavior and different jobs, such as walking and aiming. If your game allows separate settings, test each deliberately rather than assuming one value fits both. There is no universal recommended pair of numbers.

Our preview uses a radial calculation with a temporary chosen center and a rescaled output outside the selected deadzone. A game may use a different shape, range, curve or processing order. Copying the same displayed percentage does not guarantee the same response, and the preview never changes the game.

Connect DualSense to the computer or supported phone running the tester, not just to the console. Confirm detection and axes on the PS5 input page, then run the homepage resting check with the sticks released. Results describe browser-visible input, not a test performed inside the PS5.

Unwanted right-stick input is one possible cause, but game settings and mappings can also affect the symptom. Compare another game and an untouched resting run before guessing which part is faulty. A reset or larger deadzone is not proof that a worn stick was repaired.

Use a compatible USB-C data cable or a supported Bluetooth connection. For Bluetooth, follow the official pairing guidance linked on our PS5 setup page. Return to the browser, focus the page and press a controller button. Operating-system pairing does not guarantee every browser feature is available.

Charging proves power is reaching the controller, not that data is passing. Try a known data-capable cable and port, check operating-system detection, focus the browser and press a button. Also check for unsupported or blocked Gamepad API access before assuming a hardware failure.

Sony documents reset as a troubleshooting action. That is not a promise of rewritten stick calibration or repaired sensing components. A connection can recover while an off-center input remains. Hardware-writing calibration utilities are a different category and can carry risks.

Check that specific game’s controller or aiming settings and its documentation. Some games offer separate stick controls and others do not. Do not assume there is one PS5-wide menu that applies the same deadzone to every game.

Do not assume their controller-level features or profile options are identical. Check current official documentation for the exact model as well as the game’s controls. A standard browser mapping does not reveal or verify every model-specific feature.

This tester uses browser-exposed buttons and axes, not a complete DualSense hardware interface. It does not test adaptive-trigger resistance, touch coordinates, haptics, motion or audio. A touchpad click may appear as an extra button without exposing touch movement.

Connect it to the device running this site, confirm inputs on the PS4 tester and verify the axis mapping. Release both sticks and run the homepage resting capture. Repeat under the same conditions and compare the original game symptom; there is no universal pass/fail offset.

Record whether the camera, character or menu is moving and whether the sticks are released. Compare game settings, button states and resting stick input. A connection issue and an unwanted stick value need different checks, and the browser cannot diagnose a particular internal part.

DualShock 4 uses a compatible micro-USB connection. Choose a cable that supports data, not only charging, with the appropriate connector for your computer. If it charges but sends no inputs, try a known working data cable and another suitable port.

On a supported device, disconnect USB and turn off the controller. Hold SHARE and PS until the light bar flashes, then select it in the host device’s Bluetooth settings. Our PS4 setup page links Sony’s instructions. After pairing, return to the browser and press a button to check detection.

Our website cannot write hardware calibration. Some separate third-party browser utilities can change supported controller data through interfaces such as WebHID. Their risks are different from a read-only test: the calibration utility discussed in our guide warns that adjustments may worsen a problem and may not be reversible.

Software can translate a controller into a generic or virtual mapping. Our PlayStation labels are a display preference used only for standard browser mappings, not proof of device identity. Nonstandard inputs remain numbered; remapping software can change what appears.

The operating system and browser can expose different capabilities. Focus the page, press a button, check the selected device and try a supported browser and data-capable connection. If remapping software is active, check whether the browser sees a virtual device instead.

Yes. A remapper can alter the axes, buttons or virtual devices the browser receives. Stop the current test, confirm the intended device and mapping, then record a new baseline. Keep the software change noted rather than comparing the runs as identical conditions.

Use a supported USB or wireless connection to the device running this page. Confirm the controller and axes on the Xbox input tester, then run the homepage resting check without touching the sticks. Keep the connection and mapping consistent for repeated runs. Not every Xbox controller supports Bluetooth.

Check the current Xbox calibration support page and Xbox Accessories guidance for your exact model. Microsoft’s 2024 announcement confirms the tool’s existence, not a complete current eligibility list. We could not verify the detailed support page through our text-fetch tool, so we do not claim support for every Xbox One, Elite or third-party controller.

No. Recalibration concerns the relationship between stick position and interpreted input on a supported device. A deadzone ignores part of the input in a game or software layer. Their availability, persistence and effects differ; our separate preview applies neither to your controller or games.

It cannot restore physically worn material or a damaged centering mechanism. Microsoft’s recalibration announcement explicitly excludes drift due to normal wear and tear from problems the tool can resolve. A changed reading after adjustment is not proof that the hardware was repaired.

Repeated off-center stick input can cause unwanted movement, but game settings, remapping and other inputs also need checking. Record the symptom, confirm the intended device and compare a released-stick capture. Use the Xbox One checklist for older-model limits rather than applying a newer model’s procedure blindly.

A Windows connection does not guarantee that the browser has exposed the right gamepad. Focus this page, press a button, check the device selector and confirm the supported connection. If all inputs are missing, investigate connection and remapping before treating it as a stick problem.

The exact model may not support Bluetooth, or the wireless pairing, host or software path may differ. A successful wired check narrows the investigation but does not identify a failed component. Confirm the model and host instructions; Xbox Wireless and Bluetooth are not the same connection method.

Yes. Save an untouched-stick baseline, record the original settings, make one supported change and repeat with the same connection and axes. Then check the original problem in your game. Do not compare a corrected browser preview against an uncorrected resting result as if they measured the same thing.

Check operating-system connection first, then focus the page and press a controller button. Use HTTPS or localhost in a browser with Gamepad API support. Try a known data cable where supported and confirm the selected device. Unsupported access is not proof that the controller is faulty.

Browsers may wait for interaction with a connected controller before exposing it to a page. Focus the tester, press a button and choose Check connection. Detection can happen before live observation starts, but this site does not accumulate test readings until you start.

Yes. A power connection can charge a controller without carrying its input data. Try a known data-capable cable, suitable port and the correct host connection. Charging alone does not establish that a browser can read the controller.

Steam and the browser can use different input paths, mappings and permissions. Steam support does not guarantee Gamepad API exposure. Check the physical and any virtual devices, browser support and connection; do not assume a blank browser display means the controller is broken.

A remapper may translate physical inputs, create a virtual gamepad or change which axes and buttons reach the browser. The displayed values can therefore differ from another application. Confirm the intended device and mapping before interpreting a test or comparing sessions.

One possibility is that both a physical controller and a software-created virtual controller are exposed. Other devices can also appear as gamepads. Inspect the selector and check which entry responds as expected instead of assuming the extra entry is a second physical controller.

Use the device selector, then move a stick or press a button to confirm the selected entry responds. Do this before a resting capture. Switching devices or mappings interrupts an active run so readings from different configurations are not combined.

The page is responsive, but actual readings depend on the phone, operating system, browser, connection and controller. A current browser must expose the Gamepad API in a secure context. A mobile layout is not evidence of physical compatibility; this prototype’s controller/platform QA remains pending.

Start the button tester and press and release each button deliberately. Watch its current state and analog value where supplied, then check the observed-press history. A recorded press is not proof that every short transition was captured, so repeat any uncertain check.

It means the current session has not seen that input pressed. You may not have pressed it, or the browser may not expose it as expected. It does not mean failed. Clear history to repeat the checklist; current pressed state and observed history are different indicators.

A supported analog trigger can report how far it is pressed as well as a pressed state. Some devices or software paths expose only digital behavior or a different mapping. This page shows the browser’s values; it does not measure trigger force or adaptive resistance.

It may not have been fully pressed, or mapping, driver processing or the controller may affect the reported range. Compare repeated deliberate presses and the exact configuration before deciding something is damaged. This tester does not provide a universal trigger-health threshold.

No. Inputs are observed during browser animation frames, so a short press and release between observations can be missed. The tester is useful for deliberate press/release checks, not a guarantee of capturing every physical switch event.

The browser may report a nonstandard mapping, or expose extra inputs without a reliable familiar label. Numbered controls avoid guessing. Choosing the PS5, PS4 or Xbox page changes the preferred standard-layout labels; it does not identify or remap the physical hardware.

On a supported mapping, pressing a stick inward reports a button, while moving it reports axis values. Check both actions separately. If the mapping is nonstandard, identify the relevant numbered input rather than assuming a particular label or index.

No. Browser observations and display timing are not a direct measurement of the controller’s hardware polling rate or end-to-end input latency. The site shows states, axes and resting offsets; it does not produce certified timing or responsiveness scores.

Yes. Magnetic sensing avoids a sliding electrical sensing contact, but the complete stick still has mechanics, calibration and processing. The sensor label alone does not guarantee zero offset or lifelong immunity to unwanted input. Our TMR versus Hall effect guide is an explanation, not a durability benchmark.

No. An offset does not identify its cause. Centering mechanics, tolerances, software and calibration can all matter. Check repeated observations and model-specific support guidance rather than assuming that a hardware-writing adjustment is necessary or safe.

It may change behavior if contamination contributes, but it cannot reverse worn material. This site cannot determine whether cleaning is appropriate from a reading. Do not pour liquid into a controller or use an unverified internal procedure; check manufacturer guidance and warranty options first.

The action may have changed a temporary condition without addressing the cause, or different settings and test conditions may be involved. A reset can restore a connection without repairing a stick. Save repeatable before-and-after observations rather than treating a brief improvement as a permanent repair.

Yes. Browser-visible input can be checked while the controller remains assembled. Release the sticks, record a resting capture and compare game behavior. This is a non-destructive observation, not a way to inspect internal parts or certify that no fault exists.

Record the exact model, affected input, connection, game settings and repeated resting results. Check deliberate button/trigger behavior too. Compare support, warranty and repair options with the problem you can reproduce; one screenshot or unexplained percentage is not a sufficient hardware diagnosis.

Yes. Check the current terms and manufacturer support route for your model and region before internal work. Opening or modifying a controller introduces repair risks, and this site does not provide a model-specific teardown or legal determination about warranty coverage.

No. A circularity test concerns the path or shape of a stick sweep, while a resting drift check concerns input with the stick released. Different mappings and processing can change edge shapes. This site does not use a circularity score as a hardware-health verdict.