Orientation Test

Introduction

The DeviceHub Orientation Test is a free online live viewer for deviceorientation alpha, beta, and gamma alongside screen.orientation type, without OEM AR toolkits, without equating screen rotation lock with IMU angles, and without skipping iOS permission honesty. Responsive engineers debugging layout after rotate, WebXR learners reading Euler-like browser angles, educators contrasting screen.orientation with deviceorientation, QA validating gesture-gated iOS access, and support staff capturing tilt UI failures all use this HTTPS page. Many mobile sensor and advanced touch diagnostics are mobile-oriented; desktop browsers often report unsupported, and that outcome is still useful capability evidence rather than a DeviceHub defect. On iOS, DeviceOrientation and DeviceMotion access typically require a user gesture plus an explicit system permission prompt before samples flow. Compass heading derived from deviceorientation absolute alpha is approximate and suffers magnetic interference from cases, mounts, speakers, and indoor steel. The Generic Sensor API family (Accelerometer, Gyroscope, Magnetometer) is mostly Chromium-oriented, requires a secure context, and often needs an allowing Permissions-Policy, Safari and Firefox frequently omit constructors entirely. Pair with Compass Test for heading focus, Gyroscope and Accelerometer tests for rates and linear acceleration, Magnetometer Test for Chromium field vectors, Touchscreen Test when rotate changes hit targets, and Device Information for hub geometry. DeviceHub typed permissions remain none (PermissionKind has no geolocation or motion member); OS or browser prompts may still appear for geolocation or motion, editorial copy explains those prompts without inventing union values. DeviceHub does not upload orientation streams for advertising on view and will not pretend desktop always exposes full deviceorientation equal to phones. Design systems that swap navigation chrome in landscape can validate deviceorientation and screen.orientation together here before blaming CSS media queries alone. VR-adjacent web demos that need a permission-first onboarding screen can rehearse iOS gesture gates using Orientation Test as the canonical example for the whole motion family. QA on foldables should open, fold, and rotate while watching both angle sets because hinge postures create surprising screen.orientation sequences.

What this tool does

Orientation Test subscribes to deviceorientation when permitted, shows alpha/beta/gamma, reads screen.orientation.type and angle when available, and separates permission-denied from unsupported from live states so tickets gain precise language. It does not draw a full 3D digital twin, does not replace Compass Test’s heading UX copy, and does not integrate gyro rates, those remain on Gyroscope Test. Cross-links keep the suite navigable. Secure HTTPS. Local session only. Guidance explains OS rotation lock causing screen.orientation to stay portrait while your hand holds the phone sideways. Refresh after permission changes. Desktop may show only screen orientation without rich deviceorientation, still useful. Notes remind users that absolute compass aspects and magnetic interference are detailed on Compass Test rather than duplicated verbosely here. Live readouts should update frequently enough to feel responsive without claiming gyroscope-rate fidelity, point rate hunters to Gyroscope Test. When permission is denied, copy should tell users how to reopen site settings rather than only saying failed. When desktop shows screen.orientation only, the page should frame that as partial success for layout bugs, not as a broken IMU promise.

When to use it

Use Orientation Test for rotate-to-landscape bugs, iOS permission drills, teaching screen.orientation versus deviceorientation, preflight of tilt-controlled UIs, and matrices that include desktop partial support. Prefer Compass Test when the user-facing question is north heading. Prefer Gyroscope Test for angular rates. Prefer Accelerometer Test for shake and gravity. Prefer Touchscreen Test when taps fail after rotate. Prefer GPS Test for coordinates. Avoid navigation-grade attitude claims. Retest after unlocking OS rotation, after WebView switches, and after clearing site permissions. Use in workshops that demonstrate the iOS gesture gate before any alpha numbers appear. Use during responsive audits, during iOS permission UX reviews, during foldable testing, and during workshops comparing Compass absolute alpha with relative alpha. Skip when you only need maxTouchPoints, use Touchscreen Test. Skip when you need lat/long, use GPS Test. Include when tilt controls unlock features behind a consent button you are designing.

How it works

The deviceorientation event supplies alpha, beta, and gamma according to browser sensor fusion; screen.orientation supplies the screen’s type and angle relative to the viewport pipeline. iOS gates deviceorientation behind user gesture and permission. DeviceHub typed permissions remain none (PermissionKind has no geolocation or motion member); OS or browser prompts may still appear for geolocation or motion, editorial copy explains those prompts without inventing union values. Desktop browsers may expose screen.orientation while omitting rich deviceorientation, document both. Magnetic absolute heading caveats belong primarily on Compass Test. Generic Sensor Chromium paths on sibling pages are separate from these events. No DeviceHub upload on view. Emulators fake angles, label them. Secure context matters for modern hosts. Cross-origin restrictions rarely apply the same way as Generic Sensor policy, but embedded WebViews may still deny orientation. Thermal throttling can reduce event frequency. Combining orientation with multi-touch after rotate is a common PWA bug class, point to Touchscreen Test and Multi Touch Test for input confirmation. Some browsers fire orientation events at different frequencies in low-power mode; demos should disable battery saver for consistency. Screen locking to portrait in OS settings will freeze screen.orientation.type regardless of how creatively you twist your wrist, teach that before IMU debugging. Absolute versus relative alpha debates should move to Compass Test once heading is the real topic. DeviceHub keeps listeners on the page lifetime only; navigating away must remove them to avoid surprising battery use if a browser bug retained them, defensive client code matters even when content is editorial.

Step-by-step instructions

1. Open Orientation Test on a phone over HTTPS for full deviceorientation coverage.
2. Enable via user gesture and allow iOS DeviceOrientation permission if prompted.
3. Tilt through portrait and landscape; note alpha/beta/gamma and screen.orientation type together.
4. Unlock OS rotation if screen orientation type never changes despite physically rotating.
5. Compare Compass Test, Gyroscope Test, and Accelerometer Test for heading and motion without mixing field definitions.
6. Screenshot with permission state and browser name; revoke orientation permission afterward on shared devices.

Common problems

iOS frozen angles without gesture and permission. OS rotation lock confusing screen.orientation. Desktop limited data filed as total failure. Confusing screen.orientation with IMU alpha. WebView blocks. Expecting compass survey accuracy here. DevTools fake angles in RMAs. Assuming typed permissions none means no iOS prompt. Mixing gyro rate tickets into orientation without reading the right tool. Magnetic interference blamed on Orientation Test instead of Compass Test. Thick cases not relevant here but users still mention them, refocus on permission and lock state. Filing bugs for locked rotation. Expecting desktop parity. Using DevTools sensors panel secretly. Mixing up beta and gamma names in tickets. Forgetting iOS gesture. Comparing angles across browsers without timestamps. Believing Orientation Test replaces Magnetometer. Holding phones in portrait while asking why landscape CSS did not run, check screen.orientation type on the page first.

Privacy explanation

Orientation reveals how you hold and move the device. Values display locally and are not uploaded by DeviceHub for advertising on view. DeviceHub typed permissions remain none (PermissionKind has no geolocation or motion member); OS or browser prompts may still appear for geolocation or motion, editorial copy explains those prompts without inventing union values. Revoke OS permission after classroom use. Close the tab to stop event listeners. Be careful publishing orientation recordings together with GPS Test coordinates. Shared demo phones should clear live angle readouts before the next visitor. Compared with NFC payloads, orientation is continuous, but both deserve minimize-sharing habits on public forums. Continuous orientation sampling can approximate activity recognition. Keep demos short on shared hardware. Revoke permissions after class. Avoid uploading lecture captures that show a volunteer’s live angles beside their face without consent. When used with GPS Test in research, document combined sensitivity.