📷 Camera Test

🎮 Related Tools

Pair your camera test with our full suite of browser-based hardware and performance tools — all free, all private, zero installation required.

🔒 This AI-powered webcam test runs entirely in your browser. No video or audio is uploaded or stored.

🔍 What Does This Camera Test Check?

This tool uses your browser's WebRTC API to directly access your webcam and perform a full diagnostic. It measures your camera's live FPS (frames per second) using requestAnimationFrame, reads the actual resolution from the camera's track settings, and lets you record video in WebM format using the MediaRecorder API. Need to test your microphone too? Try our online microphone test.

🤖 Test Webcam with AI Face Detection

Our AI webcam test includes real-time face detection powered by TensorFlow.js and face-api.js — running entirely in your browser with no data sent to any server. Once your camera is active, the AI engine detects your face position, draws landmark overlays, and provides live feedback on centering, eye openness, and smile detection. It can simultaneously track hand positions and body pose, making it one of the most comprehensive camera performance tests online. Toggle individual AI features — face detect, landmarks, hands, background segmentation, and pose skeleton — on or off at any time without interrupting the live stream.

📊 Check Webcam FPS and Real-Time Metrics

Accurate webcam FPS testing online requires measuring directly from the live video stream — not from a static property. This tool tracks frames in real time using requestAnimationFrame and updates your FPS display continuously. You can monitor resolution, frame rate, and camera latency simultaneously. Whether you're troubleshooting low-FPS issues or benchmarking a new webcam, these real-time webcam metrics give you the ground truth your system delivers — not a spec-sheet estimate. For a standalone frame-rate benchmark, see our an FPS test tool.

🎬 Record and Download Webcam Video

The built-in recording studio lets you record webcam video online free using the browser's MediaRecorder API — no plugins, no extensions, no uploads. Click Record to start capturing, click Stop to finish, and the file downloads instantly to your device in WebM or MP4 format (Safari). You can also play back the recording in the browser before saving. This webcam test with recording capability is ideal for checking video quality, lighting, and audio sync before important calls. For audio-only testing, use our microphone test.

🛠️ Troubleshooting: Webcam Not Working?

If your webcam isn't working or shows a black screen, here are the most common fixes:

• Camera permission denied: Click the lock icon in your browser's address bar and set Camera to "Allow", then refresh the page.
• Webcam black screen fix: Another app (Zoom, Teams, OBS) may have exclusive access to your camera. Close those apps and try again.
• Low FPS webcam fix: Low FPS is often caused by poor lighting, a USB hub with insufficient bandwidth, or outdated camera drivers. Try connecting directly to a USB 3.0 port and ensure drivers are current.
• HTTPS required: Browser camera access requires a secure (HTTPS) connection. If you're on HTTP, your browser will block camera access.
• Outdated browser: Use the latest version of Chrome, Firefox, Edge, or Safari for full WebRTC and AI feature support. You can also verify your hardware inputs with our keyboard test.

🔒 Is My Camera Data Private?

Absolutely. All processing happens entirely in your browser. No video frames, screenshots, or any camera data are ever uploaded to a server. Files you capture download directly to your device. This tool requires HTTPS to function, ensuring your browser session is encrypted. The same privacy guarantee applies to our microphone test.

How a Webcam Turns Light into Video

Every number this camera test shows is the end result of a chain: light passes through a lens, lands on a sensor, is processed inside the camera, travels over USB, is decoded by your computer and handed to the browser, then painted on your screen. When something looks wrong, knowing which link is responsible saves a lot of guesswork.

The Parts Inside a Webcam

A typical USB webcam is a small circuit board with four main parts: an image sensor, a lens in front of it, an image signal processor (ISP), and a USB controller that talks to your computer. Laptop cameras use the same parts, often connected internally over USB.

The image sensor

The sensor is a grid of millions of light-sensitive sites, usually a CMOS chip. Each site collects light during the exposure and converts it into an electrical charge, which is then read out and turned into a number. A color filter array (most commonly a Bayer pattern of red, green and blue filters) sits on top, so each site only records one color; the full-color image is reconstructed later by interpolating between neighboring sites. That reconstruction step, called demosaicing, is one reason fine patterns such as striped shirts can show color fringes on cheaper cameras.

Webcam sensors are physically tiny, so each pixel catches relatively little light. That is the root cause of the grainy, smeared look many webcams have indoors: with less light per pixel, random variation in the signal (noise) is large compared with the signal itself. A high megapixel figure on a very small sensor does not guarantee a cleaner picture.

The lens and field of view

The lens focuses the scene onto the sensor and sets the field of view, which is how much of the room the camera sees. Webcam lenses are usually fixed in focal length, so the angle of view cannot change optically. Wide lenses show more of the room but make faces smaller and can bend straight lines near the edges. Most webcams also have a fixed aperture, which is why they rely so heavily on exposure time and gain to cope with dim rooms.

The image signal processor

The raw sensor data is not a usable picture. The ISP, a dedicated chip or block inside the camera, turns it into one. It handles demosaicing, automatic exposure, automatic white balance, noise reduction, sharpening, gamma and color correction, and often lens shading correction (brightening the corners that a small lens leaves darker). Effects such as the picture slowly brightening after you switch on a lamp, or a face looking waxy in low light, are decisions the ISP makes in real time. Your browser receives the ISP's output, not the sensor's raw data, so any tool running in a browser measures the processed image.

How Video Gets from the Camera to the Browser

Most external webcams follow the USB Video Class (UVC) standard. UVC is the reason a webcam usually works on Windows, macOS and Linux without installing a vendor driver: the operating system already knows how to talk to any UVC device. During setup, the camera tells the computer which formats, resolutions and frame rates it supports, and the computer picks one.

Uncompressed, MJPEG and H.264 formats

A UVC camera typically offers some combination of three kinds of format:

• Uncompressed YUV (often listed as YUY2 or YUYV). Each pixel is sent as brightness plus shared color information, at 16 bits per pixel. The picture is free of compression artifacts and needs no decoding, but it uses a lot of bandwidth.
• MJPEG (Motion JPEG). Each frame is compressed individually as a JPEG image inside the camera. This cuts bandwidth dramatically, which is why it is the usual format for 1080p on USB 2.0 cameras. The cost is some compression detail loss and extra work for your computer, which has to decode every frame.
• H.264. Some cameras can encode video themselves using a codec that compresses across frames as well as within them. It needs even less bandwidth than MJPEG, but support varies between cameras and operating systems, and many current webcams do not offer it.

The operating system and browser negotiate the format with the camera. It matters because it affects the available resolution and frame rate, how much processor time decoding takes, and whether fast movement shows blocky artifacts.

USB 2.0 vs USB 3.0 bandwidth

USB 2.0 is rated at 480 megabits per second, and the share a single video device can actually use is noticeably lower than that headline figure once protocol overhead and scheduling are taken into account. USB 3.0 (also sold as USB 3.2 Gen 1) is rated at 5 gigabits per second. A little arithmetic shows why the difference matters for uncompressed video:

This is why a USB 2.0 webcam that is happily delivering 30 fps at 1080p must be sending compressed frames, and why the same camera may only offer uncompressed high resolutions at very low frame rates. It also explains a classic symptom: if the system ends up negotiating an uncompressed mode for some reason, a camera that should be smooth can fall to single-digit frame rates at HD resolutions. Sharing one USB controller between a webcam, an external drive and an audio interface can create a similar squeeze, because they all draw from the same pool of bandwidth.

Resolution, Aspect Ratio and Capture Modes

Resolution is the number of pixels in each frame, written as width × height. Aspect ratio is the shape of the frame, the ratio of width to height. The two are connected but not the same: 1280 × 720 and 1920 × 1080 are different resolutions with the same 16:9 shape, while 640 × 480 is a 4:3 shape that looks squarer.

Common webcam modes

A camera's "capture modes" are the specific combinations of format, resolution and frame rate it offers. A 1080p webcam might, for example, offer 1080p at 30 fps in MJPEG, 720p at 60 fps in MJPEG, and 640 × 480 at 30 fps uncompressed. Higher resolution captures more detail, which helps when you are small in the frame or when video will be cropped, but it also means more data to move, decode and process every second.

What the browser reports vs what the camera can do

A web page cannot open a camera at "maximum quality" directly. It asks for what it would like, and the browser works out what to deliver. When you click the start button, this camera test asks for an ideal size of 1920 × 1080 with no audio. "Ideal" means "as close to this as you can", not "exactly this", so the browser chooses the nearest mode the camera and system can provide. The resolution shown afterwards is read back from the live video track, which is what you actually received.

Two consequences follow. First, a 4K webcam will normally show 1920 × 1080 here, because that is what the page requested; that is expected and does not mean the camera is faulty. Second, the delivered resolution is not always a native mode of the sensor. Some browsers can crop or scale frames to satisfy a request, so a camera might be capturing at one size internally while the page receives another. The Resolution Test in the Tools tab, described below, is a way of probing which sizes the camera will accept.

Why the browser may pick a lower mode

If the reading is lower than you expected, common reasons include:

• The camera simply does not offer a higher mode in a format your system will use. Many built-in laptop cameras are 720p.
• Bandwidth limits, for instance a USB 2.0 connection, a shared hub, or a long or low-quality extension cable, so only lower modes are offered at usable frame rates.
• Another application or browser tab already has the camera open at a particular mode. On some systems a second consumer of the same camera receives whatever the first one negotiated.
• Virtual camera software sits between the physical camera and the browser and outputs its own fixed resolution.

Frame Rate Explained

Frame rate, measured in frames per second (FPS), is how many separate images are captured or shown each second. Video calls commonly run at around 30 fps, and 60 fps looks noticeably smoother on fast motion. Below roughly 15 fps, movement starts to look choppy to most people.

Camera frame rate vs display refresh rate

Two different "rates" get mixed up constantly. The camera frame rate is how often the camera delivers a new image. The display refresh rate is how often your monitor redraws the screen, commonly 60 Hz on ordinary screens and 120 Hz, 144 Hz or more on high-refresh gaming monitors and some laptops and phones. Browsers schedule their own drawing to match the display, and a video element simply shows the latest camera frame each time the screen refreshes. A 30 fps camera on a 60 Hz screen therefore shows each camera frame for about two refreshes. The screen updating 60 times a second does not make the camera capture 60 images a second, and a fast monitor cannot make a slow camera smoother.

Why FPS drops in low light

At 30 fps each frame has at most about 33 milliseconds to be exposed. In a bright room that is plenty. In a dim room, the camera's automatic exposure wants more light per frame than 33 ms can collect. It has two options: raise the gain (amplify the signal, which also amplifies noise) or lengthen the exposure. Lengthening the exposure beyond the frame interval is only possible by producing fewer frames, so many webcams quietly drop to around 15 fps or lower when the room gets dark. Some cameras call this "low-light compensation" in their settings software. This is the single most common reason a webcam that is smooth at midday is sluggish in the evening, and it is why improving lighting often fixes "low FPS" without touching any other setting.

Longer exposures also cause motion blur: anything that moves while the shutter is effectively open gets smeared. If your hand looks like a ghostly streak when you wave it, the camera is using a long exposure.

Bandwidth, CPU load and background tabs

Beyond light, the frame rate you actually see can be limited by:

• Link bandwidth, as described above.
• Processor load. Decoding MJPEG frames, scaling them, running effects and painting the page all use CPU or GPU time. When the computer is busy with other work, frames can be dropped before they reach the screen. Optional extras on this page, such as the AI detection overlays, add their own load.
• Browser throttling. Browsers slow down or pause drawing work in tabs that are hidden or minimized, to save power. Measurements based on the page's drawing loop pause too, which is why this page stops updating its counters while you are in another tab. Always read the numbers with the tab visible and in the foreground.

Latency: Where the Delay Comes From

Latency is the delay between something happening in front of the lens and it appearing on screen. A local preview usually feels close to instant, but there is always some delay, built up from several stages:

1. Exposure. Each frame takes time to collect light, longer in dim conditions.
2. Sensor readout and ISP processing inside the camera, plus any compression the camera performs.
3. Transfer over USB to the computer.
4. Decoding and buffering by the operating system and browser.
5. Waiting for the next screen refresh, since a new frame can only appear when the display next redraws.
6. The display itself, which adds its own processing delay.

In a video call, network transmission and the other person's decoding add more on top. If your preview here feels responsive but people on calls see you lagging, the bottleneck is likely the network or the calling app rather than the webcam.

Exposure, White Balance, Focus, Dynamic Range and Noise

These five properties decide most of what people mean by "good" or "bad" webcam picture quality. Each one has a recognizable look in a live preview.

Exposure

Exposure is how bright the camera makes the image, controlled by exposure time and gain. Webcams adjust it automatically from the brightness of the whole frame or its center, which is why a bright window behind you turns your face into a silhouette. Watch for white areas with no detail (clipped highlights) or a picture that keeps brightening and darkening, which means automatic exposure is hunting.

White balance

White balance is how the camera decides what "neutral white" is, so colors look natural under different lights. Warm indoor bulbs are orange compared with daylight; automatic white balance tries to cancel that tint. When it gets it wrong, skin looks orange, gray or slightly green, and a white wall is visibly tinted. Mixed lighting, such as a window on one side and a warm lamp on the other, is especially hard, and a large area of one strong color in the frame can also push automatic white balance off.

Focus: fixed focus vs autofocus

Many webcams, especially built-in laptop cameras, are fixed focus: the lens is set so that everything from roughly arm's length to the far wall is acceptably sharp, and nothing very close will ever be crisp. Autofocus webcams move the lens to focus on whatever they judge to be the subject. Autofocus can give a sharper face at desk distance but may briefly blur the picture as it refocuses when you move. A uniformly soft image at a sensible distance usually points to a dirty lens, a protective film left on it, or a low-resolution stream.

Dynamic range

Dynamic range is the span between the darkest and brightest parts of a scene that the camera can record with detail at the same time. Small webcam sensors have limited dynamic range, so a scene with both deep shadows and a bright window forces a compromise: either the shadows go black or the window goes white. Some cameras offer a "backlight compensation" or HDR mode that lifts shadows at the cost of a flatter look; more even lighting usually helps more.

Noise

Noise is the random speckle and color blotches visible in darker areas of the picture. It rises sharply as light falls, because the camera raises gain to compensate. Many ISPs then apply strong noise reduction, which hides speckle but also smears fine detail, giving skin and hair a painted, waxy appearance. Noise is also harder to compress, so a noisy picture can look blockier in a call. Adding light in front of you is the most effective way to reduce it.

Reading the Camera Test Results, Panel by Panel

Some readings come straight from what the browser reports about the camera; others are calculated by the page from the pixels of the live video. Reported values describe the stream's configuration, while calculated values describe the scene, so they change when you move, change the lighting or point the camera elsewhere.

Frame Rate, FPS Trend and Max FPS

The large Frame Rate number in the Live Metrics panel, and the FPS badge on the preview, are calculated by counting how many times the browser's animation loop runs, updated about twice a second. The browser runs this loop once per display refresh while the page is visible and able to keep up. In practice, that makes the counter a measure of how smoothly your browser is drawing the page while the camera preview is running.

That has some practical implications:

• On a typical 60 Hz screen, a healthy system tends to show a number near 60, even if the webcam itself is delivering 30 frames per second. On a high-refresh display, the number can be higher.
• If the counter falls well below your screen's refresh rate, the browser is struggling to keep up. Common causes are heavy CPU or GPU load, the AI detection features, or many other busy tabs.

For the frame rate the camera stream itself is configured for, open the Device Info panel and look at the Frame Rate line, which is read from the video track's settings. This is the rate the browser and camera agreed on, such as 30.0 fps. It is a configured value, so in very low light a camera that lengthens its exposure may deliver fewer frames than this line says. To spot that, wave a hand slowly and watch for stepping or streaking in the preview.

Max FPS in the View tab's Camera Information list mirrors the live counter rather than a hardware maximum. The FPS Trend graph plots recent values of the same counter, so you can see whether it is steady or dips, for example when you switch the AI features on or open another heavy application. When AI detection is active, the page shows a note that a slight FPS drop is normal; it also adapts how often the AI models run when the frame rate falls, so the AI work does not overwhelm slower machines.

Camera Info: Name, Resolution, Megapixels and Aspect Ratio

The Name is the label your operating system gives the camera, for example the model name of a USB webcam or a generic name for a built-in camera. It confirms the browser opened the camera you meant to test. If the wrong one opened, change the camera in your browser's site settings, or disconnect the one you are not testing, and start again.

Resolution is the width and height of the frames you are receiving, read from the track settings. Megapixels is simply width × height divided by one million, so 1920 × 1080 shows 2.07 MP and 1280 × 720 shows 0.92 MP. This is the size of the video frames, not the camera's still-photo rating, which on phones and some webcams is much higher. Aspect Ratio is the width and height reduced to their simplest whole-number ratio. Standard modes give familiar answers such as 16:9 or 4:3. An unusual ratio, such as 30:17, usually means the browser or camera has cropped or scaled to a non-standard size, which is worth noting if you see unexpected cropping in calls.

Facing, FOV Class, Stream Type and Image Mode

Facing shows the direction the camera reports: "user" for a front or selfie camera, "environment" for a rear camera, or N/A when the camera does not report a direction, which is normal for most desktop webcams.

FOV Class is a rough label based on the frame's megapixels and facing direction; for a front or desktop camera, 8 MP or more is labelled Wide, 2 MP or more Standard, 0.5 MP or more Narrow, and below that VGA. It does not measure the lens angle, which browsers do not expose, so treat it as a description of the stream size rather than of the optics. The Field of View line in the Real-Time Metrics panel is a separate estimate based only on the aspect ratio, so 16:9 shows as "Wide (16:9+)" and 4:3 as "Standard". To judge the real field of view, sit where you normally sit and look at how much of the room is visible.

Stream Type and Image Mode are labels worked out from the facing direction and the device name. Desktop webcams usually show "Local Camera" and "Standard". If Image Mode shows "Infrared" or "Depth", the browser may have opened the wrong sensor of a multi-sensor (face-login) camera.

Latency, Smoothness and the Latency Test

The Latency reading in Live Metrics, and the Avg, Min, Max and Jitter boxes in the Latency panels, are built from per-frame timing information the browser provides for video frames. Where the browser supplies a capture timestamp for each frame, the page measures the time between that timestamp and the moment the frame is scheduled to be shown. Where it does not, the live meter falls back to the time between successive frames being presented, which for a steady 30 fps camera would sit around 33 ms and mostly reflects the frame interval rather than true delay. Browsers without per-frame callbacks fall back to a simple timer, which says little about the camera. The live figures keep a rolling window of the latest 60 samples.

The Latency Test in the Tools tab is the more careful measurement. It collects 60 frames that carry a capture timestamp and reports:

• Avg: the mean delay across the sample.
• Min and Max: the quickest and slowest frames.
• Jitter: the standard deviation, which shows how much the delay varies from frame to frame. Low jitter means a steady pipeline; high jitter means frames arrive unevenly, which looks like small stutters.

The result message is colored green when the average is under 50 ms, amber from 50 to 100 ms and red above that. The test needs per-frame timing support, which the page notes is available in Chrome and Edge; if the browser does not provide capture timestamps, the test reports that there were not enough frames. The capture timestamp is assigned by the computer, not inside the camera, so exposure and processing time in the webcam and your monitor's delay are not included. The figure is best used for comparisons, such as two USB ports, or AI features on and off.

Smoothness is easy to misread. It is not a frame-rate figure. Twice a second the page takes a tiny grayscale copy of the frame and compares it with the previous one; the less the picture changes, the higher the percentage. A still scene reads close to 100%, and waving your arms drops it sharply. Read while you sit still, it helps spot a shaky mount, a flickering light or exposure that keeps pumping.

Color and Image Metrics

The Color & Image panel describes the overall color and brightness of the scene, using a small, downscaled copy of the frame sampled twice a second.

• Avg RGB is the average color of the whole frame, shown as a swatch and a hex code. In a neutral room under good white balance it tends to be a muted gray-beige. A strongly orange, blue or green swatch in a room you know to be neutral suggests a white balance problem.
• Brightness is the average perceived brightness as a percentage of maximum, using standard weights that count green most and blue least, as human vision does.
• Hue, Saturation and Lightness describe that same average color in HSL terms. Hue is the color angle (0° red, around 120° green, around 240° blue), saturation is how strong the tint is, and lightness is how light it is. A low saturation figure means the average color is close to gray, which is what you would expect from a balanced picture of an ordinary room.
• Luminosity is the perceived brightness of the average color, again as a percentage.
• Num Colors counts the distinct colors in the small sample after rounding each channel to 32 levels. Because the sample is only 80 × 45 pixels, the count can never exceed 3,600. Dark or flat images give low counts; noise also adds "colors", so compare it only within the same scene.

All of these depend on what is in front of the camera: a red wall gives a red average and says nothing about the camera. For color judgements, include your face and a neutral background.

Device and Stream Details

The Device & Stream panel collects details that are mostly descriptive:

• Microphone and Speaker show whether the browser can see audio devices. The camera itself is opened without audio; the microphone is only requested if you start a recording. For a full audio check, use the microphone test.
• Video Std is a label based only on the frame rate: rates near 25 or 50 fps show "PAL", rates near 15, 30 or 60 fps show "NTSC", and others show "Unknown". Webcams are not broadcast devices, so this only tells you which family of frame rates the stream uses.
• Bitrate (shown as "Bitrate (est.)" in the View tab) is an estimate, not a measurement. It multiplies width × height × frame rate by a fixed factor of 0.07 bits per pixel, a rough figure for compressed webcam video. At 1280 × 720 and 30 fps that gives about 1.9 Mbps; at 1920 × 1080 and 30 fps, about 4.4 Mbps. It is a ballpark, not what any calling app will actually send.
• PNG Size and JPEG Size are the sizes the current full-resolution frame would have if saved as a PNG and as a high-quality JPEG. They update every few seconds. Detailed, sharp, well-lit frames and noisy frames both produce larger files, while dark, blurry or flat frames produce smaller ones. If the image gets brighter when you add light but the JPEG gets smaller, noise has probably dropped.
• Quality shows the most recent overall score from the Video Call Readiness test or the Quality Analyzer, whichever you have run, as a score out of 100 with a word rating.

The Device Info panel on the right repeats the key values in one block, which makes it the easiest place to copy your camera's configuration when reporting a problem to someone helping you.

Live Quality Bars and Real-Time Metrics

The Live Quality panel, and the matching bars in the View tab and the Real-Time Metrics panel, are recalculated from the live picture a little more than once a second, using a copy of the frame up to 320 pixels wide. Each bar runs from 0 to 100%, and the color shifts from green to amber to red as the score falls.

Sharpness

Sharpness is estimated by measuring how much the brightness changes sharply between neighboring pixels (a standard edge-detection measure called the variance of the Laplacian), shown on a logarithmic scale. The caveat is that it depends on the scene: a plain, evenly lit wall has few edges and scores low even if the camera is perfectly focused, while a bookshelf full of titles scores high. Use it to compare the same scene before and after cleaning the lens, moving closer or farther, or improving the light. The Sharpness Trend line plots recent values and is colored green when sharpness is rising, red when it is falling, and blue when it is flat, which is handy for watching an autofocus camera settle.

Lighting and Exposure

The Lighting bar scores the average brightness of the frame, with the highest scores for a mid-range average and lower scores for very dark or very bright pictures. The Exposure label in Real-Time Metrics turns the same average into words: Underexposed for very dark frames, Slightly Bright or Overexposed for bright ones, and Good in between. Because it is a whole-frame average, a well-exposed face against a dark room can still read as dim, and a dark face against a bright window can read as fine. Check your face in the preview as well.

Noise

Noise is estimated by measuring the variation inside 20 small patches spread across the frame and taking a low value from that set, on the reasoning that the flattest patches show sensor noise rather than real detail. A higher percentage on this bar means a cleaner picture. Very textured scenes, such as a patterned wallpaper, can make an otherwise clean camera look noisier, so compare like with like.

Stability

Stability uses the same frame-to-frame difference as Smoothness, scaled differently: an unchanging picture scores 100%, and the score falls as more of the picture changes between samples. Real-Time Metrics calls this Motion Jitter. Sitting still in front of a camera on a stable mount should give a high reading. A low reading while you are still suggests a wobbling mount, a vibrating desk, flickering light or exposure that keeps changing.

Contrast

Contrast is the difference between the darkest and the brightest pixel in the sampled frame. It shows whether the picture uses the full tonal range, but because it only looks at the extremes, a single bright lamp or one dark corner can push it to 100%. A low reading means everything is a similar shade, typical of a washed-out or very dim picture.

Color Temp and Color Balance

The Color Temp bar compares the average red level with the average blue level and labels the picture Warm, Cool or Neutral. The Color Balance line in Real-Time Metrics shows the same label. Strongly colored backgrounds or clothing will move it, so treat a Warm or Cool label as a prompt to look at your skin tones in the preview, not as a fault in itself.

Analysis Tab: Advanced Diagnostics and the Quality Analyzer

The Advanced Diagnostics grid in the Analysis tab runs a second set of pixel tests about once a second on a 640 × 360 copy of the frame and turns them into labels:

For Auto Exposure, "Adjusting" right after a lighting change is normal. "Fluctuating" when nothing in the room changes suggests exposure is hunting, often due to a flickering light or a very contrasty scene.

The face-related cells (Face Detect, Eyes Detect, Smile Detect, Blink Rate, Face Distance, Face Alignment and Hands Detected) rely on a face detector. The page uses the browser's built-in detector where one is available, which it notes is limited to Chrome and Edge, and otherwise shows "Chrome/Edge" in those cells while the pixel tests keep running. When you switch on AI Detection and it finds a face, the AI results take over those cells.

The Quality Analyzer on the right of the Analysis tab gives a one-shot report from a single frame. It scores four things and combines them with these weights:

The overall result is shown as a letter grade: A for 85 and above, B from 70, C from 55, D from 40, and F below that. Short comments explain what pulled the grade down, such as the scene being too dark, the image being soft, or high motion. Because it uses one frame, run it a few times while sitting still.

The Video Call Readiness Score

The Video Call Readiness test combines seven factors into a score out of 100, shown as a ring with a label: Call Ready from 85, Good from 70, Needs Work from 40, and Not Ready below 40. The same score fills the Quality Rating ring in the right-hand panel. The weights are:

Two details make it worth running the test more than once. First, the lighting and face-size readings are smoothed across runs, so one odd frame does not swing the result. Second, lighting and distance suggestions only appear after the problem has shown up in several consecutive runs, which stops the list from flickering with advice. Centering advice appears straight away. For Face Framing, the score falls if your face fills too little of the frame (you are far away) or too much (you are very close), or if it sits well off-center. If no face is found, the factor scores low and the page suggests turning on AI Detection, which is more reliable in poor light or at an angle than the browser's built-in detector.

Resolution Test Results

The Resolution Test in the Tools tab uses the camera that is already running. For each of four sizes, 1920 × 1080, 1280 × 720, 640 × 480 and 320 × 240, it asks the browser to switch the live track to exactly that size, then reads back what it actually got. Each row shows a tick with "Supported · got W×H", or a cross with "Not supported". At the end, the page restores the preview's original size.

Read it this way:

• A tick where the "got" size matches the row is the clearest result: the stream can be delivered at that size.
• Because some browsers are able to scale video to meet a request, a tick does not always prove the sensor has a matching native mode. It does prove that apps in this browser can receive video at that size.
• A cross on 1920 × 1080 only is typical of a 720p camera, or a 1080p camera limited by bandwidth or another app.
• The test does not check sizes above 1080p, because the page's camera request is capped at 1080p.

What "Normal" Generally Looks Like

There are no hard pass/fail limits, but these ranges are a reasonable guide for an ordinary webcam in a reasonably lit room, with AI features off:

Patterns matter more than single readings: low sharpness plus high noise plus a dim exposure label all point to lighting; a low frame rate counter with good image readings points to processing load; a low configured frame rate with dim readings points to the camera lengthening its exposure.

Settings on This Page That Do and Don't Affect the Readings

Some of the page's own controls change what you see without changing what is measured, and it helps to know which:

• Image Filters (presets and the brightness, contrast, saturation, hue, blur and sepia sliders) are applied to the on-screen preview only. The quality calculations read the camera's original frames, so filters do not improve or worsen the scores. Screenshots, however, do include the active filter and mirror setting, so reset filters before saving a frame you want to judge for color.
• Mirror / View only flips the preview.
• Zoom uses the camera's own zoom if the camera reports that it supports it; otherwise it enlarges the preview on screen. Hardware zoom changes what the camera captures and therefore the readings; on-screen zoom does not.
• AI Detection does not alter the camera's picture, but it uses processing power, so it can lower the frame rate counter and affect latency readings on slower computers. Take your baseline readings with it off.

A Thorough Webcam Check, Step by Step

This routine uses the page's features in an order that separates causes from each other. Doing it once in full gives you a reference to compare against later.

Before You Start

• Close other apps that might be using the camera, including calling apps running in the background and other browser tabs with camera access, so this page gets the camera to itself.
• Plug an external webcam directly into the computer rather than through a hub or extension cable for the first test. You can repeat the test on the hub later to compare.
• If the camera has a privacy shutter or a protective film over the lens, open or remove it. Wipe the lens gently with a clean microfiber cloth; fingerprints are one of the most common causes of a hazy, low-contrast image.
• Set up your normal call lighting, plus a lamp you can switch on and off.

The Routine

1. Start the camera and confirm the basics

Start the camera and allow access. Wait about ten seconds for the camera's automatic exposure and white balance to settle; many webcams look too dark or tinted for the first moments. Then check the Camera Info panel: is the Name the camera you meant to test, and is the Resolution what you expected? Open the Device Info panel and note the configured Frame Rate.

2. Probe the supported sizes

Open the Tools tab and run the Resolution Test. Note which of the four sizes get a tick and what size each one actually delivered. If 1920 × 1080 fails on a camera sold as 1080p, try a different USB port, ideally a USB 3 port on the computer itself, and run the test again. If it now passes, the earlier connection was the limit.

3. Take a baseline with AI off

Sit still at your normal distance for twenty seconds with AI Detection switched off. Watch the Frame Rate counter and the FPS Trend graph: is the counter steady and close to your screen's refresh rate? Look at the Smoothness and Stability readings: they should be high while you are still. Glance at Auto Exposure in the Analysis tab: it should settle to Stable. Write down the frame rate counter, the configured frame rate, and the Stability percentage.

4. Test the lighting response

Switch your extra lamp off, wait ten seconds, then switch it back on and wait again. In each state, note the Brightness percentage, the Exposure label, the Noise bar and the configured Frame Rate in Device Info. You are looking for three things:

• How much noise rises as the room gets darker.
• Whether the configured frame rate drops, or motion looks steppy when you wave a hand, in the dark. That means the camera trades frame rate for exposure.
• How long Auto Exposure takes to return to Stable. A few seconds is common.

5. Check focus and sharpness

Hold a page of printed text or a book cover at the distance where your face normally is, then move it slowly closer to the camera. Watch the Sharpness bar and the Sharpness Trend line. On an autofocus camera, sharpness should dip briefly and recover as the lens refocuses. On a fixed-focus camera, sharpness will fall steadily once the page gets closer than the lens's near limit; note roughly where that happens so you know how close you can sit. Then put the page away and check the Analysis tab's Sharpness label on your normal scene.

6. Check white balance and color

Make sure Image Filters are set to None. With your face and some neutral background in view, look at the Avg RGB swatch, the Color Temp label and the Color Balance cell. If the label says Warm or Cool, check your skin in the preview: does it look natural to you? Try turning off one light source if you have mixed lighting, wait for the camera to settle, and compare. A white sheet of paper held up to the camera is a good reference: it should look white or light gray, not orange or blue.

7. Measure latency

In the Tools tab, run the Latency Test while sitting still, and note the average and the jitter. Then switch on AI Detection with Face Detection and Face Mesh enabled and run it again. If the average or jitter rises a lot with AI on, the delay is coming from processing load on your computer rather than from the camera. If you want to compare USB ports or hubs, repeat the test on each with the AI features off.

8. Take a screenshot and a short recording

Click 📷 Screenshot to save a full-resolution JPEG of the current frame. Open it at full size and look for smeared detail from noise reduction or a soft area that suggests a smudged lens. Then record a ten-second clip, talking and moving your hands naturally. Play it back and watch for stutters in movement and for whether your speech stays in step with your lips. The recording stays in your browser until you download it.

9. Run the readiness and quality reports

Finally, run the Quality Analyzer two or three times, then run the Video Call Readiness test three or four times while sitting in your normal call position. Note the overall score and the lowest-scoring factor, which is usually the one most worth improving first.

Matching Symptoms to Likely Causes

Combinations of readings narrow down the cause more reliably than any single number.

Keeping a Record for Later Comparisons

Webcam problems are often intermittent, so a short record makes it easier to spot what changed. After your first full check, note:

• The camera name, resolution and configured frame rate from Device Info.
• Which sizes passed the Resolution Test, and on which port.
• The Latency Test average and jitter, with AI off.
• The Quality Analyzer grade and the Video Call Readiness score, plus the lowest factor.
• A screenshot taken from your usual seat with filters off.

Later, repeat only the steps related to the symptom and compare. If the numbers match your record but calls still look poor, look at the calling app or your network instead of the camera. The same habit works for the rest of your setup; the device tools collection covers your other inputs too.

🩺 Troubleshooting by Error Message and Symptom

Webcam problems fall into two groups: either the browser never gets a video stream, and this page shows an error card in the preview area, or the stream starts but the picture is wrong. Once you know which layer is refusing (the browser, the operating system, another app or the hardware), the fix is usually quick.

📋 Reading the Error Card and Status Bar

When you click "▶ Test My Camera", the page asks the browser for a video stream. If the browser refuses, it returns a named error, and the page translates that name into one of a few cards shown over the preview. The status bar under the buttons repeats a short version of the same message. The table shows what each card means in plain terms.

Any other failure appears only in the status bar, as "Camera error:" followed by the browser's own wording. That wording comes from the browser itself, so it is worth searching for exactly as written.

Keep in mind that Retry simply asks the browser again, so change a setting before you press it. Also, a card reports the browser's verdict, not always the real cause; if the obvious fix does nothing, try the fixes for the neighboring cards too.

🔒 "Camera Permission Denied"

This card means something said no before the camera was ever opened. There are three places that can say no, and you need all three to say yes.

1. The website permission in your browser. Every browser keeps a per-site list of allowed and blocked sites. If you clicked Block or Don't Allow once, many browsers remember it and stop asking. That is why the prompt sometimes never appears again.
2. The operating system's privacy setting for the browser app. Windows, macOS, Android, iOS and ChromeOS all let you deny camera access to a whole application. When the browser itself is blocked, every website in it fails, and the browser's own site settings may still say "Allow".
3. Policy set by someone else. Work and school devices can be configured so the camera is blocked for all sites, or allowed only for a list of approved ones. Parental controls can do the same on family devices.

The prompt never appeared

If you clicked the button and went straight to the error card without seeing a question, the decision was already stored. Look for a small camera icon with a slash or a red mark in the address bar; clicking it usually offers a way to reset the choice. If there is no such icon, open the site information panel to the left of the address and look for Camera there. After changing it, reload the page rather than pressing Retry, because some browsers only apply a changed permission to a fresh page load.

You allowed it, but it still fails

When the browser shows Camera as allowed for this site and you still get the card, the block is almost always at the operating system level. Jump to the section for your system below and look for the setting that controls camera access for apps, and specifically for desktop apps on Windows. On a Mac, a browser that has just been granted camera access in System Settings often needs to be quit completely and reopened before the change takes effect.

🚫 "Camera Already in Use"

This card appears when access was granted but the device itself refused to start. On Windows this is the classic sign that another program is holding the camera, because many webcam drivers on Windows let only one application read the sensor at a time. The second app either gets an error or, in some cases, a black picture.

The first step is obvious, and the page already suggests it: close the other video apps. What catches people out is everything that keeps running after you think you closed it.

• Apps that minimize to the tray. Many meeting and chat apps keep running in the notification area or menu bar after you close their window. Quit them from the tray icon or the app menu, not just the window's close button.
• Other browser tabs. A meeting tab, another webcam test or a page with a video widget can hold the camera. In most browsers, the tab that is using the camera shows a small camera or red dot icon on its tab or in the address bar.
• Background helpers. Webcam vendor utilities, streaming software, security camera apps and some face sign-in features can open the camera without a visible window.

If you cannot find the culprit, a restart releases every handle on the device. If the card still appears straight after a fresh restart with nothing else open, treat it as a driver or hardware issue and skip ahead to the hardware checks.

On macOS, several apps can usually share one camera, so this card is rarer there. When it does appear on a Mac, a stuck camera process or a faulty USB connection is a more likely cause than a competing app.

📷 "No Camera Found"

Here the browser looked at the list of devices and found no video input at all. The page does not show a camera picker, so this message reflects what the browser itself can see. Work from the outside in:

1. Is the camera powered and connected? For an external webcam, unplug it, wait a few seconds and plug it into a different USB port directly on the computer. Many webcams have a small activity light that blinks briefly when connected.
2. Is there a hardware switch? Some laptops have a physical camera kill switch on the side or a function key that disconnects the camera electrically. When it is off, the camera disappears from the system entirely, which produces this card rather than a black picture.
3. Can the operating system see it? Check Device Manager on Windows, System Information on a Mac or the list of video devices on Linux (all covered below). If the OS cannot see it, the browser cannot either.
4. Is it disabled? On Windows, a device can be disabled in Device Manager or turned off on the Cameras settings page. On some laptops the firmware setup screen (the BIOS or UEFI menu) also has an option to disable the integrated camera.

The page listens for devices being plugged in, so after you connect a camera you can usually click Retry without reloading. If the camera only shows up after a reload, that is also fine and not a sign of a fault.

⬛ Black Screen with No Error

A black preview with no error card means the browser opened a camera and is receiving frames, but the frames are dark. That narrows things down considerably, because permissions and drivers are clearly working at least partly. Check these in order:

• Privacy shutter or lens cover. Many laptops and webcams have a small sliding cover over the lens. It is easy to close by accident and easy to miss, especially on dark bezels. A slider that shows a red or orange mark usually means closed.
• Keyboard camera toggle. On some laptops the camera function key cuts the image instead of removing the device, which leaves you with a black or gray frame and sometimes an on-screen icon showing a crossed-out camera.
• A very dark room. In near darkness, many webcams produce a frame that looks black. Turn on a light facing you and watch the Lighting reading in the live metrics. If it rises when you add light, the camera is fine.
• The wrong sensor. Laptops with face sign-in often have an infrared camera next to the normal one. If the browser opens that one, you may see a nearly black or dim grayscale image. The Image Mode line in the Device Info panel reads Infrared when the camera's name identifies it as one.
• Another app holding the device. On some systems a second app gets black frames instead of an error. Closing other video apps, as described above, is worth trying even when no card appears.

A black picture that turns into a normal image after a second or two is usually just auto-exposure settling, which is normal for most webcams when they first start.

🧊 Frozen or Stuck Image

A picture that stops moving can come from the camera, the connection or the page view itself. Before you troubleshoot hardware, rule out the page-level causes:

• You are watching a recording, not the live feed. After you stop a recording, the main preview plays back what you recorded so you can check it. That can look like a looping or stuck picture. Click "📡 Back to Live" to return to the camera.
• You are looking at a screenshot. A captured screenshot appears as a still image below the controls. It is easy to mistake for the preview on a small screen.
• The tab was in the background. Browsers slow down or pause work in hidden tabs to save power. When you come back, the live readings may take a moment to catch up. Switching away from the tab also stops a recording made with the main Record button.

If none of those apply and the status bar says "Camera was disconnected.", the stream ended underneath the page. That usually means the USB connection dropped, the system put the port to sleep, or the operating system revoked access mid-session (for example, because you toggled a privacy switch). Reconnect and click "▶ Test My Camera" again.

A freeze with no message at all points at the camera or its connection: a loose cable, a hub short of power or bandwidth, or USB power saving, all covered in the hardware section below.

🔀 The Wrong Camera Opens

This page does not offer its own camera picker. It asks for "a camera" and lets the browser decide which one, so the camera you get is the browser's default. If you have a laptop with a built-in camera plus a USB webcam, or software that installs virtual cameras, the default may not be the one you want to test. You can confirm which device you got after the feed starts: the Label line in the Device Info panel and the Camera Name field under Camera Information both show the device's name.

To change which camera opens:

• Chrome and Edge: open the browser's camera settings page (chrome://settings/content/camera or edge://settings/content/camera, typed into the address bar) and pick your camera from the drop-down at the top. That choice becomes the default for sites that do not request a specific device.
• Firefox: the permission prompt includes a list of cameras. If Firefox stopped asking because you ticked "Remember this decision", remove the saved permission for the site so the prompt with the list comes back.
• Any browser: unplugging or disabling the camera you do not want is the bluntest method, but it is definitive and useful for testing.
• Phones and tablets: the browser usually opens the front-facing camera for a page like this one. The Facing line in the Device Info panel shows which one you got: "user" means front and "environment" means rear.

🐢 Very Low or Uneven FPS

Choppy video has two very different sources: the camera delivering fewer frames, or the computer struggling to draw and analyze them. The page gives you enough information to tell which.

Check what the camera is set to

Open the Device Info panel and read the Frame Rate line. That figure is the frame rate the browser reports for the camera track. If it says something like 5, 10 or 15 fps, the camera itself is running slowly, and no amount of closing tabs will help. If it says 30 or 60 fps but the picture still stutters, the bottleneck is more likely the computer or the browser.

Dim light is a frequent cause

Most webcams lengthen their exposure in dim light to gather more of it. Each frame then takes longer, so the frame rate drops, often to half or less of the normal rate. It is a very common reason a webcam that "should be 30 fps" runs at 15. Turn on a lamp facing you, or move to a window, and watch whether the frame rate recovers within a few seconds. If it does, the camera is fine and your room is the issue. Some webcam control apps include a setting, often called low-light compensation or auto-exposure priority, that allows this slowdown; turning it off keeps the frame rate steady at the cost of a darker image.

The AI features use a lot of processing

Face detection, landmarks, hand tracking, pose and background segmentation all run on your own device. With several of them turned on, a slower laptop or phone can spend most of its time on analysis. The page shows a notice when AI detection is active to remind you that a slight FPS drop is expected. To measure the camera's real smoothness, switch AI detection off and compare.

The rest of the system

• Battery saver and power modes. Laptops on battery saver, and phones in low-power mode, limit processor speed and can reduce how often the browser updates the page.
• Other heavy tabs and apps. Games, video editors and many open browser tabs compete for the same processor.
• USB bandwidth. Several cameras, capture devices or drives sharing one hub or controller can starve a webcam of bandwidth, which often shows up as a low frame rate at high resolutions.
• Resolution. Some webcams can deliver their full resolution only at a lower frame rate. The Resolution Test in the Tools tab checks which of four common sizes your camera accepts, which helps when you suspect this.

🔄 Mirrored or Upside-Down Image

Mirroring is a display choice, not a fault

Many video apps show your own preview flipped left-to-right, like a mirror, because that is how people are used to seeing themselves. Raise your right hand and it appears on the right side of the screen, just as it would in a bathroom mirror. The other people on a call are typically sent the unflipped picture, so text on your shirt or a sign behind you reads correctly for them even though it looks backwards to you. Apps differ in their defaults, and many let you switch the self-view mirroring on or off; changing that option affects only what you see.

On this page the main preview starts in Natural View, meaning it is not flipped. You can switch between Mirror View and Natural View from the "🔄 Mirror / View" button below the preview. If text looks backwards here with Natural View selected, the camera or its driver is flipping the image, and that is worth looking into in the camera's own software.

Upside down usually means a driver or mounting issue

A picture that is upside down (rotated 180 degrees) is never a normal mirroring choice. The usual causes are:

• A webcam mounted upside down on a clip or stand. Some cameras correct for this automatically; many do not.
• A rotation or flip option set in the webcam's own control software, sometimes left over from a previous setup.
• A generic or incorrect driver on a laptop whose camera module is physically installed upside down, which the correct driver would normally compensate for. This is more common on Linux and on Windows after a clean install without the manufacturer's drivers.

The Mirror Tool in the Tools tab can flip its own view horizontally, vertically or both, which shows what a corrected picture would look like. It only changes the display on this page, so the lasting fix belongs in the camera software or driver.

💡 Flickering or Rolling Bands Under Indoor Lights

If you see horizontal dark bands that roll slowly up or down the picture, or a brightness pulse that comes and goes, the cause is almost always the room lighting rather than the camera.

Why it happens

Mains electricity alternates at either 50 Hz or 60 Hz depending on the country. Many lamps, especially fluorescent tubes and some LED bulbs, get brighter and dimmer with each half of that cycle, so they actually pulse 100 or 120 times per second. Your eyes do not notice, but a camera can. Most webcams use a rolling shutter that reads the sensor one row at a time from top to bottom, so when the light pulses partway through a frame, some rows are captured brighter than others. The result is bands. If the camera's frame rate does not line up exactly with the flicker, the bands drift, which is why they appear to roll.

How to fix it

1. Set the anti-flicker option to match your mains frequency. Many webcams have a setting called anti-flicker, power line frequency or something similar, with 50 Hz, 60 Hz and sometimes Auto. It lives in the webcam maker's app, in some built-in camera apps, or on Linux in the camera's controls. Choose the frequency used where you live. Much of Europe, Africa, Asia and Australia uses 50 Hz; North America and parts of South America use 60 Hz; Japan uses both, depending on the region. If you are unsure, try both and keep the one that removes the bands. A browser page cannot change this setting, so it must be done outside the browser.
2. Change the light. Daylight, incandescent bulbs and good-quality LED lamps flicker much less. Switching off one problem light, often a ceiling fluorescent, can fix it on the spot.
3. Avoid low dimmer settings. Many dimmable LEDs flicker more at low brightness. Turning the dimmer up, or off the dimmer entirely, often helps.

Flicker that appears as random flashes rather than steady bands, or that continues in daylight, is a different problem. It points toward auto-exposure hunting, a loose connection or a failing camera rather than the lights.

🎨 Green, Purple or Other Color Casts

Color problems fall into two families: a believable picture that is simply the wrong shade, and a picture that is obviously broken. They have different causes.

The whole image is tinted but otherwise normal

This is usually white balance. The camera guesses what "white" should look like from the scene, and strong colored light misleads it. A bright monitor glow, a colored wall, a lamp with a strong tint or mixed daylight and indoor light can all push the image toward blue, orange, green or magenta. Try these:

• Put a neutral light source in front of you and reduce colored light from behind or beside you.
• Give the camera a few seconds after changing the lighting; automatic white balance adjusts gradually.
• If your webcam software allows manual white balance, set it with a white sheet of paper in view, or switch back to automatic if someone left it on a fixed value.
• Watch the Color Temp and Color Balance readings on this page while you change the lighting. They give you a consistent reference instead of relying on your eyes alone.

The image is bright green, pink or purple, or broken into blocks

A solid green frame, a strong magenta cast with strange colors, or green blocks and streaks are not lighting issues. They usually mean the video data is being decoded wrongly or arriving damaged. Typical causes are a driver problem, a faulty or overlong USB cable, a weak hub, or a camera that is failing. Try a different port and cable, update or reinstall the driver, and test the same camera on another computer. If the fault follows the camera to a second device, the camera is the problem.

A dim purple or gray picture

An image that looks washed-out purple or grayscale in a normal room is often an infrared camera, as described under black screens above. In daylight, a strong purple or pink cast on an otherwise normal camera can also mean that its infrared filter is damaged or missing, which lets invisible light affect the colors. That is a hardware fault and cannot be fixed in software.

Check the page's own filters

Before you blame the camera, make sure no color filter is active on this page. The Image Filters popup includes Warm, Cool, Invert and a Hue slider, any of which can produce an unnatural color. "↺ Reset All Filters" returns everything to normal.

💻 Fixes by Operating System and Browser

The steps below use the setting names that have been stable for several releases, but menus get renamed and rearranged from time to time, so names may vary by version. If a path does not match exactly, searching the settings app for "camera" almost always finds the right page.

🪟 Windows 10 and Windows 11

Camera privacy settings

Windows has its own on/off switches for the camera that sit above every browser setting. If any of them is off, browsers fail even with the site allowed.

• Windows 11: open Settings, then Privacy & security, then Camera. Make sure Camera access is on, Let apps access your camera is on, and Let desktop apps access your camera is on.
• Windows 10: open Settings, then Privacy, then Camera. Make sure access for this device is on, Allow apps to access your camera is on, and Allow desktop apps to access your camera is on.

The desktop apps switch is the one people miss. Chrome, Firefox, Edge and other common browsers are desktop apps rather than Store apps, so they are controlled by that last switch, not by the list of individual Store apps above it.

Windows 11 also has a Cameras page under Settings, then Bluetooth & devices. It lists connected cameras and lets you disable or re-enable one. On some systems it also offers basic image controls such as brightness and contrast, and a preview you can use to confirm the camera works outside the browser.

Device Manager basics

Device Manager shows whether Windows recognizes the camera and whether its driver loaded. Right-click the Start button and choose Device Manager, then expand Cameras. Some webcams appear under Imaging devices instead, and a few under Sound, video and game controllers.

• A down-arrow on the icon means the device is disabled. Right-click it and choose Enable device.
• A yellow warning triangle means Windows found the device but the driver has a problem. Double-click it and read the Device status box for a short explanation.
• No camera listed at all, with the camera connected, means Windows cannot see the hardware. Choose View, then Show hidden devices, to check whether it was detected earlier and is now missing; then try another port and cable.

To reinstall the driver, right-click the camera, choose Uninstall device, and then either restart or choose Action, then Scan for hardware changes. Windows reinstalls a driver automatically. If the problem started after an update, the Driver tab in the device's properties may offer Roll Back Driver, which returns to the previous version.

A quick check outside the browser

Windows includes a Camera app. If it shows a normal picture while browsers fail, the hardware and driver are fine and the problem is a permission or browser setting. If the Camera app also fails, focus on the camera, driver and Windows settings first. Close the Camera app before testing in the browser again, so it does not hold the device.

🍎 macOS

On recent versions of macOS, open System Settings, then Privacy & Security, then Camera. Each app that has asked for the camera appears with a switch. Turn on the switch for the browser you are using. Older versions use System Preferences, then Security & Privacy, then the Privacy tab, then Camera, with checkboxes instead of switches.

A few points specific to Macs:

• Restart the browser after changing the setting. macOS often asks you to quit and reopen the app. If it does not ask but the camera still fails, do it anyway.
• A browser missing from the list has never requested the camera. Load this page in that browser and click "▶ Test My Camera"; macOS should then ask and add the browser to the list.
• The green light next to the built-in camera turns on whenever the camera is active. If it is on while nothing should be using the camera, some app is still holding it.
• Managed Macs from work or school can have camera access controlled by a profile. In that case the switch may be grayed out, and only the administrator can change it.

To confirm the camera works outside the browser, open an app that uses it, such as Photo Booth or FaceTime. If that works and the browser still does not, the problem is the browser's setting or permission.

🐧 Linux

Linux distributions differ, so these checks are deliberately general. Most USB webcams follow a standard class (UVC) that the kernel supports without extra drivers.

• Check that the kernel sees the camera. Cameras appear as video device files, such as /dev/video0. Many webcams create two entries, one for the picture and one for metadata, so seeing two for one camera is normal. If the v4l-utils package is installed, v4l2-ctl --list-devices shows which entries belong to which camera.
• Check permissions on the device. On most desktop distributions, the user logged in at the screen gets access automatically. On some systems, especially minimal installs, your user must belong to the video group. Logging out and back in is required after changing group membership.
• Check sandboxed browser packages. Browsers installed as Snap or Flatpak packages run with their own permission rules, which can block the camera even when the system allows it. Your software center or the package's command-line tool usually has a permissions or interfaces view for the camera.
• Check camera controls. v4l2-ctl --list-ctrls lists the camera's settings, including brightness, white balance and, on many cameras, power_line_frequency, which is the anti-flicker setting discussed earlier. Graphical tools for the same controls exist in most distributions' repositories.

💼 ChromeOS

Chromebooks handle the camera mostly through the browser, with a system-wide switch on top in recent versions.

• System switch: in Settings, look under Privacy and security for a camera access control (in some versions it sits inside a Privacy controls section). When it is off, no app or site can use the camera.
• Site permission: works the same as in Chrome on other systems, through the site information panel or the browser's camera settings page.
• Built-in Camera app: open it from the launcher. If it shows a picture, the hardware is fine.
• Managed Chromebooks: school and work devices often have camera access set by the administrator, with a building icon or a "managed by" note next to the setting.

Some Chromebooks have a physical privacy switch or shutter for the camera; check the bezel and sides if the picture is black.

🤖 Android

Two settings matter on Android: the browser's site permission and the system's app permission for the browser.

• Site permission in Chrome: tap the icon to the left of the address, then Permissions, or go to the three-dot menu, then Settings, then Site settings, then Camera. Make sure this site is not in the blocked list.
• App permission: open the system Settings app, then Apps, find your browser, then Permissions, then Camera, and choose Allow only while using the app or Ask every time.
• Quick settings toggle: recent Android versions add a camera access tile in the quick settings panel. When it is off, every app gets a blank or blocked camera. Swipe down from the top of the screen to check.

Phone makers sometimes rename or move these menus. If the phone's own Camera app works but the browser does not, the problem is one of the permissions above.

📱 iPhone and iPad

On iPhone and iPad, all browsers use the same underlying engine, so the steps are similar whichever browser you use.

• Safari, for this site: tap the page settings button in the address bar (shown as "aA" or a similar icon), choose Website Settings, and set Camera to Ask or Allow.
• Safari, for all sites: open the Settings app, then Safari, and find Camera under the settings for websites. If it is set to Deny, no site can ask.
• Other browsers: open the Settings app, scroll to the browser's name and make sure its Camera switch is on.
• Screen Time: if the camera is restricted under Screen Time's content and privacy restrictions, it is blocked for every app. This is common on children's devices and on devices set up by an organization.

Open the page in a full browser rather than an app's built-in link viewer, which may not support camera access. If the camera stops after you switch apps or lock the screen, return to the page and click "▶ Test My Camera" again.

🌐 Chrome and Microsoft Edge

Chrome and Edge share the same engine, so their camera handling is nearly identical.

• For this site: click the icon at the left of the address bar. Depending on the version, it looks like a padlock or a small settings icon. Find Camera and choose Allow, or use Reset permission to make the browser ask again.
• For all sites: type chrome://settings/content/camera in Chrome or edge://settings/content/camera in Edge. This page shows the default behavior, lists blocked and allowed sites, and has the drop-down for choosing which camera sites get by default.
• Address bar camera icon: when a site is using the camera, or has been blocked from it, a camera icon appears at the right of the address bar. Clicking it shows the current state and lets you change it.

🦊 Firefox

• For this site: click the permissions icon at the left of the address bar. A crossed-out camera means the site is blocked; click the X next to the blocked permission to remove it, then reload.
• For all sites: open Settings, then Privacy & Security, scroll to Permissions, and click the Settings button next to Camera. You can remove saved decisions here, and you can also tick the option to block new requests, which explains a prompt that never appears.
• Choosing a camera: Firefox's permission prompt has a list of available cameras. If you have more than one, check that the right one is selected before clicking Allow.

🧭 Safari on Mac

• For this site: with the page open, choose Safari in the menu bar, then Settings for this website (the wording varies slightly by version), and set Camera to Ask or Allow.
• For all sites: open Safari's Settings (called Preferences in older versions), go to the Websites tab, and select Camera in the left column. You can set a default for other websites and change any saved decision.

Safari also asks the system for camera access, so the macOS privacy setting applies too.

🔐 Why HTTPS Is Required

Browsers only allow camera access on pages they consider secure, which means pages served over HTTPS, plus local development addresses such as localhost. On a plain HTTP page, the camera API is simply not offered, so no amount of permission changes will help. The rule exists so that a network eavesdropper cannot inject code into an unencrypted page and use it to watch you.

This page handles that case for you. If it is ever opened over HTTP, it shows the Insecure Connection card and a banner, then redirects to the HTTPS address after a three-second countdown, with a link to go immediately. If you keep ending up on HTTP, an old bookmark, a proxy or a network filter may be rewriting the address; update the bookmark to the https:// version.

🧩 Extensions, Security Software and Managed Devices

Browser extensions

Privacy and security extensions can block camera access, and some do so silently. Others, such as virtual background or meeting helper extensions, can intercept the camera. The quickest test is a private window (Incognito in Chrome, InPrivate in Edge, a Private Window in Firefox or Safari). Most extensions are disabled there by default. If the camera works in a private window and not in a normal one, an extension is the likely cause. Disable your extensions one at a time to find it. Note that a private window starts with no saved permissions, so you will be asked to allow the camera again.

Security software

Some antivirus and internet security suites include webcam protection that blocks applications they do not recognize. When that happens, the browser may report a permission error or a device that cannot be opened. Look in the security software's privacy or webcam section for a list of allowed apps, and add your browser if it is missing.

Work, school and enterprise policies

Organizations can control camera access through device management. In Chrome and Edge, typing chrome://policy or edge://policy shows any policies applied to the browser; policies with names such as VideoCaptureAllowed or VideoCaptureAllowedUrls control the camera. In Firefox, about:policies shows the equivalent. On Windows, Group Policy or management software can also turn off the camera for the whole device. If a policy is blocking the camera, settings you change yourself will either be grayed out or reverted, and only your IT administrator can change it.

🔌 App Conflicts, Hardware Checks and When to Replace

When permissions are correct and the browser is up to date, the remaining causes are other software competing for the camera and the physical path from the lens to the computer.

🎥 Conflicts with Zoom, Teams, Meet, Discord and OBS

Exclusive access

On Windows in particular, many cameras can be opened by only one application at a time, and the first app keeps it until it lets go. Meeting apps usually release the camera when a call ends, but some keep it open while a preview window, settings page or waiting room is showing. Streaming and recording software such as OBS keeps a camera open for as long as a scene containing it exists, even when you are not streaming or recording.

A practical routine for finding the conflict:

1. End any call and close the meeting app's window, then quit it from the system tray or menu bar.
2. Close other browser tabs that might use the camera, including web-based meeting tabs. Closing the tab is more reliable than leaving the meeting inside it.
3. In streaming software, remove or hide the camera source, or quit the program completely.
4. Click Retry on the error card, or "▶ Test My Camera" again.

Doing it in the opposite direction is useful too. If this page has the camera and a meeting app then complains that the camera is busy, click "■ Stop" here first. Stopping the test releases the camera, and the camera light should go off.

Virtual cameras

Streaming tools, some meeting apps, phone-as-webcam apps and background-effect programs can install a virtual camera. It appears to the browser as one more device, with a name that usually mentions the software. A virtual camera only shows a picture when its software is running and sending one, so selecting it by accident gives you a blank screen, a logo or a frozen frame.

Virtual cameras can also become the default device. If this page keeps opening one, change the default camera in your browser settings as described in the wrong-camera section, or uninstall the virtual camera if you no longer use the software that created it.

🔍 Physical Checks Before Anything Else

Several of the most common "dead camera" reports turn out to be a switch or a cover.

• Privacy shutter. Look closely at the lens. Many built-in and external cameras have a small slider; some laptops have a shutter that is controlled by a button or key instead. A closed shutter normally gives a black picture rather than an error.
• Camera function key. Check the top row of the keyboard for a key with a camera symbol, sometimes with a slash through it. On many laptops it must be pressed together with the Fn key. Depending on the model, turning it off either blanks the picture or removes the camera from the system, which leads to "No Camera Found". Some laptops show a small on-screen message when you press it.
• Camera light. Most cameras have an indicator light that turns on while the camera is active. If the light comes on when you start the test but the picture is black, the camera is running and the problem is the image path: a shutter, a dark room, the wrong sensor or a fault. If the light never comes on, the camera is not being opened at all.
• Lens condition. A smudged or scratched lens causes a hazy or blurry picture that no setting will fix. Clean it gently with a soft, dry microfiber cloth.

🧷 USB Ports, Hubs and Cables

External webcams depend on a good USB connection for both power and data. Problems here cause many intermittent faults: freezes, disconnects, green or garbled frames and low frame rates.

• Plug directly into the computer. Unpowered hubs, monitor USB ports and keyboard pass-through ports may not supply enough power or bandwidth for a webcam. If the camera works when plugged straight into the computer, the hub was the problem.
• Try another port, preferably on the other side. On many computers, ports on different sides connect to different internal controllers. Moving the camera can get it away from another busy device.
• Avoid long passive extensions. Extra cable length can cause dropouts. If you need distance, an active (powered) extension is more reliable than chaining passive cables.
• Check the cable itself. On cameras with a detachable cable, try a different one. Some cables, especially cheap ones, are made for charging only and carry no data, so a camera connected through one may get power but never appear as a device.
• Watch for wiggling. If the picture drops or freezes when you move the cable, the cable or connector is damaged. That is a hardware fault, not a setting.

🔋 Power Saving and Sleep

Operating systems try to save power by turning off devices they think are idle. Webcams sometimes get caught by that.

• USB selective suspend on Windows. In the advanced settings of your power plan (Control Panel, then Power Options, then Change plan settings, then Change advanced power settings), the USB settings section contains USB selective suspend. Disabling it can stop a camera from dropping out after it has been idle. On some Windows 11 systems this option is hidden or managed by the manufacturer's power software.
• After sleep or hibernation. A camera that stops working after the computer wakes up, but works after a restart, is a common driver quirk. Unplugging and reconnecting an external camera usually wakes it. For a built-in one, a restart is the practical fix, and a driver update is worth checking.

🧰 Updating, Reinstalling or Rolling Back Drivers

Before changing drivers, confirm that the camera fails in more than one app. A camera that works in the system's own camera app does not have a driver problem.

When a driver is the likely cause, a sensible order is:

1. Install pending system updates. On Windows, check Windows Update, including any optional driver updates it offers. On a Mac, camera drivers come with macOS updates.
2. Get the driver from the right source. For a built-in laptop camera, use the laptop maker's support page for your exact model, because laptop cameras often rely on chipset or platform drivers too. For an external webcam, use the webcam maker's site. Avoid third-party "driver updater" tools.
3. Reinstall cleanly. In Device Manager, uninstall the camera device; if offered, tick the option to remove the driver software. Restart, and let Windows or the maker's installer set it up again.
4. Roll back if the problem started after an update. The Roll Back Driver button in the device's properties returns to the previous version. If it is grayed out, there is no previous version stored.

Many external webcams work with the generic driver built into the operating system and do not need a vendor driver at all. The maker's software is mainly useful for extra settings such as anti-flicker, white balance and image rotation. If the generic driver gives a working picture and the vendor software breaks it, uninstalling the vendor software is a reasonable choice.

🔁 Testing on Another Device to Isolate the Fault

The fastest way to tell a camera fault from a computer fault is to move one variable at a time.

For a built-in laptop camera you cannot move the camera, but you can still swap the other parts of the test. A clean user account, a different browser and the system's camera app between them rule out most software causes. If a known-good USB webcam works on the same laptop, the operating system and browser are fine and the built-in module is the likely fault.

🗑️ When to Give Up and Replace the Camera

Before you spend money, work through this checklist. If you can tick every item and the fault is still there, replacing the camera is a reasonable decision rather than a guess.

1. Shutters, function keys and side switches are all confirmed open or on.
2. Browser site permission and operating system camera access are both allowed, and no policy or security software is blocking it.
3. Every other app that could use the camera is closed, and the problem remains after a full restart.
4. The camera fails in the system's own camera app as well as in the browser.
5. For an external camera: it has been tried in at least two ports directly on the computer, with a different cable where the cable is detachable.
6. The driver has been updated, cleanly reinstalled and, if the problem started after an update, rolled back.
7. The same fault appears on a second computer (external camera), or a different webcam works fine on the same computer (built-in camera).
8. The fault is a hardware symptom: no detection on any device, constant disconnects with a good cable, persistent green or garbled frames, a pink or purple cast in normal daylight, a picture that stays blurry with a clean lens, or a mark or spot that stays in the same place in every frame.

For a built-in laptop camera, check the warranty first; otherwise a USB webcam is often simpler than a repair.

When you choose a replacement, test it here as soon as it arrives. Check that the Resolution Test accepts the sizes you need (it checks up to 1920 × 1080), that the Frame Rate line stays steady when you turn on a normal room light, and that the image does not band under your usual lighting. Problems found in the first days are much easier to return than ones found weeks later.

Lighting and Framing: Getting a Better Picture From the Camera You Already Have

Once the test confirms that your webcam works, the next question is usually "why does it still look bad?" In most home and office setups the answer has less to do with the camera than with the light falling on you and where the camera sits. A webcam sensor is tiny compared with the sensor in a dedicated camera, so it collects very little light. When light is scarce, the camera compensates automatically: it raises the gain (which adds grain), lengthens the exposure (which can blur movement and, on many models, lowers the frame rate), and smooths away detail with noise reduction. Give the same camera plenty of even light and much of that processing simply switches off. The sections below cover how to light yourself, where to put the camera, and how to use the readings on this page to check your changes rather than guessing.

Why Light Changes Image Quality So Much

Every webcam has a limited range of brightness it can record at once. Inside that range you get clean colors and visible detail; outside it, bright areas clip to flat white and dark areas sink into muddy gray noise. Good lighting is really about two things: getting enough light on your face that the camera does not have to boost its signal, and keeping the brightness of the whole scene within a range the camera can handle in one exposure.

That second point is why a bright room can still produce a poor picture. If a lamp or window behind you is far brighter than your face, the camera has to choose which one to expose for. Most webcams average the whole frame, so they darken the image to tame the bright area and your face ends up underexposed. Moving or adding a single light often fixes this more effectively than any setting change.

You can see the effect in the live readings here. The Sharpness and Noise values usually improve together when you add light, because a camera that is not straining for exposure produces a cleaner, crisper frame. If you tend to see "Image appears soft" or "Camera noise is high" in the Video Call Readiness suggestions, try more light before you assume the lens or sensor is at fault.

Key, Fill and Back Light Explained

Photographers and video crews describe lighting in terms of three roles. You do not need three lamps to benefit from the idea; often one light plus a wall or window can cover two of the roles. Knowing what each one does helps you work out what is missing when a picture looks flat, harsh or murky.

Key Light

The key light is your main source and does most of the work. Place it in front of you and slightly to one side, roughly at or a little above eye level, so it lights your whole face and puts a small, natural shadow on the far side of your nose. A key light placed straight in front of you gives the most even result and suits video calls well; moving it further to the side adds shape but also deeper shadows. A key light placed too low (for example, a laptop screen lighting you from below) creates an unflattering upward shadow, while one placed directly overhead casts shadows into your eye sockets.

Fill Light

The fill light softens the shadows the key light creates. It is usually dimmer than the key and sits on the opposite side. A fill does not have to be a lamp: a white wall, a sheet of white card, or a pale curtain on the shadow side reflects some of the key light back onto your face. The aim is to lift the darker side of your face so it still shows detail, not to remove shadows entirely, which can make the image look flat.

Back Light

A back light (also called a rim or hair light) sits behind you, out of the frame, and adds a thin edge of light to your hair and shoulders. It separates you from the background, which helps both viewers and background blur or segmentation features find the outline of your head. Keep it noticeably dimmer than the key light. If the back light is too strong or shines toward the lens, it causes the backlighting problem described further down.

Making the Most of Window Light

A window is often the best free light source you have. Daylight through glass is broad and soft, which flatters faces and produces gentle shadows. The rule is simple: face the window, or sit at an angle to it, with the camera between you and the glass. Never sit with the window behind you if you can avoid it.

• Facing the window: gives the most even light and is the easiest setup for calls. If direct sun comes through, a thin white curtain or blind diffuses it so it stops being harsh.
• Window to one side: gives more shape and depth, but the far side of your face can go quite dark. Add a fill on the shadow side, such as a white wall, a pale sheet or a dim lamp.
• Changing conditions: daylight shifts in brightness and color through the day and with passing clouds. A setup that looks good at 10 a.m. can be too dim by late afternoon.
• Direct sunlight: hard sun on your face creates bright patches and sharp shadows that the camera cannot expose well at the same time. Diffuse it, or move out of the direct beam.

Avoiding Backlight and Silhouettes

Backlighting is the most common lighting mistake on video calls. It happens whenever the brightest thing in the frame is behind you: a window, a bright lamp, a white wall lit by sun, or even a large bright monitor visible in the background. The camera exposes for that bright area, and your face drops into shadow.

Signs that you are backlit:

• Your face looks darker than the wall behind you.
• The background is bright but washed out, with little detail.
• The image gets noticeably darker when you lean away from the camera and lighter when you lean in (because your face then fills more of what the camera measures).
• The BG Light reading in the advanced diagnostics shows "Bright" or "Overlit" while your face still looks dim on screen.

Fixes, from easiest to most involved:

1. Turn around or rotate your desk so the window or lamp is in front of you instead of behind.
2. Close the blind or curtain behind you, or switch off the lamp in the background.
3. Add light to your face so it is at least as bright as the background.
4. Change the camera angle so the bright area is out of the frame.
5. As a last resort, use a backlight-compensation or exposure setting in your camera software (see the camera settings section below). This brightens your face but will usually blow out the background further.

Mixing Color Temperatures

Light sources have different colors even when they all look "white" to your eyes. Daylight from a window tends toward blue, while many household bulbs are warm and yellow-orange. Your eyes adapt to the mix without you noticing, but a camera's white balance can only correct for one color at a time. When two very different sources light your face from different sides, one side of your face can look orange and the other blue, and the auto white balance may drift back and forth as you move.

Ways to avoid mixed lighting:

• Use bulbs of the same color for all the lights that reach your face. Bulb packaging usually lists a color temperature in kelvin (K); lower numbers are warmer, higher numbers are cooler. Lamps marketed as "daylight" are closer to window light, while "warm white" bulbs suit evening-only setups.
• If you rely on a window during the day, use a cooler bulb as your extra light so the two match more closely.
• If you work mostly in the evening, switch off cool overhead light and use warm lamps only, or vice versa.
• If you have an adjustable LED panel or ring light, adjust its color until your skin looks natural in the preview rather than trusting the number on the dial.

The Color Temp reading in the live quality panel labels the overall frame as Warm, Neutral or Cool, and the Color Balance diagnostic reports Balanced, Warm (R+), Cool (B+) or Green+ by comparing the average red, green and blue across the image. Both look at the whole frame, so a brightly colored wall, a red jumper or a green plant filling the shot will push the reading even when your lighting is fine. Treat them as a nudge to look closer, and judge your skin tone by eye.

Cheap and Practical Lighting Setups

You do not need studio equipment for a noticeably better picture. The table below lists setups that use common household items, roughly from simplest to most involved.

A few general tips apply to all of these. Larger light sources give softer light, so a lamp bounced off a big wall looks gentler than the same lamp pointed straight at you. And if you wear glasses, raise the light a little or move it further to the side so its reflection lands outside your lenses; you can check this live in the preview.

Reading This Tool's Lighting Indicators

Several readings on this page react to light, and they are most useful when you watch them while you change something. Here is what each one looks at, based on how the page calculates them:

• Lighting (live quality panel and Video Call Readiness): based on the average brightness of the frame. The readiness test suggests adding light when the scene is very dim and reducing bright lights or windows behind you when the scene is heavily overexposed.
• BG Light (advanced diagnostics): samples small patches in the four corners of the frame, on the assumption that the corners usually show background, and labels them Dark, Dim, Normal, Bright or Overlit. A bright corner reading combined with a dim face is the classic backlight pattern.
• Color Temp and Color Balance: compare the red and blue (and green) content of the frame, as described above.
• Contrast: reflects how widely brightness varies across the frame. Very low contrast often means flat, foggy light or a dirty lens; very high contrast can mean harsh light and deep shadows.

These are estimates from pixel values, not calibrated light-meter readings. They cannot tell the difference between a dark sweater and a dark room, or between a white wall and an overexposed one. Use them to compare one arrangement with another in the same room, which is where they are genuinely helpful.

Camera Position and Eye Level

After lighting, camera height makes the biggest difference to how you come across. A laptop sitting flat on a desk places the lens well below your eyes, so the camera looks up at you: viewers see more of your chin and ceiling and less of your eyes. Raising the camera to roughly eye level gives a more natural, conversational view.

• Laptops: raise the whole laptop on a stand, a stack of sturdy books or a box, and use a separate keyboard and mouse if you need to type during calls.
• External webcams: clip the camera to the top edge of your monitor, then raise the monitor or the camera mount so the lens is near your eye line.
• Looking at the camera: people often look at the other person's face on screen, which makes them appear to look slightly away. In an interview or presentation, glancing at the lens when you make a key point helps it feel like eye contact.
• Distance from the screen: sitting too close to a wide-angle laptop camera exaggerates whatever is nearest the lens, usually the nose and forehead. Sitting back a little, around an arm's length, keeps proportions natural.

Headroom, Distance and Using the Grid Overlay

Framing is about where you sit inside the rectangle. Two common problems are too much empty space above your head, which makes you look small and sunken in the frame, and the top of your head being cut off, which feels cramped. For most calls, aim for a small gap between the top of your head and the top of the frame, with your eyes about a third of the way down from the top and your shoulders visible.

The Grid Overlay button under the preview offers several guides. Rule of Thirds splits the frame into a three-by-three grid; placing your eyes near the upper horizontal line is a reliable starting point. Center Cross helps you sit centered for formal calls and presentations. Diagonal and Golden Ratio lines suit more creative compositions, such as a streaming layout where you sit to one side. The grid is drawn only on this page's preview; it is not added to screenshots or recordings.

The Video Call Readiness test also checks distance and centering. It measures how much of the frame your detected face fills and suggests moving back if you are so close that your shoulders are lost, or moving closer if your face is very small. If your face sits off-center it tells you which direction to move. Run it, adjust, and run it again; it is quicker than asking someone how you look.

Angle, Background Tidiness and Depth

A camera placed straight in front of you at eye level is the safe choice. A small offset, with the camera slightly to one side and you turned a little toward it, can look more relaxed for informal calls or streams, but large angles make it obvious that you are not looking at the viewer.

Your background matters because the eye goes to anything bright, high-contrast or moving. A few practical checks:

• Remove clutter that sits right at the edges of your head, such as a lamp that seems to grow out of your shoulder or a shelf line that cuts through your neck.
• Check what is visible for privacy: whiteboards, mail, calendars, family photos and screens can all be readable on a high-resolution camera. Taking a screenshot and viewing it at full size (see the capture section below) is a good way to check.
• Avoid strong patterns and busy shelves directly behind you. They draw attention and are hard for video compression to encode cleanly, which can make your whole picture blockier on a call.

Depth makes an image look more professional. If you can, sit a few steps away from the wall behind you rather than with your back against it. This keeps your shadow off the wall, lets the background fall slightly out of focus on cameras with a limited depth of field, and gives a back light space to work. A darker background with a single lamp or plant a little way behind you often looks better than a flat, evenly lit wall.

Camera Settings, Hardware Choices and Why Calls Look Different

Lighting and position fix most problems, but sometimes you need to go further: adjust the camera's own settings, decide whether a different camera would help, or work out why the picture looks fine on this page and poor on a call. This part covers each of those in turn, in general terms that apply to most systems.

Camera Controls in Your Operating System and Companion Apps

Many webcams expose controls beyond what a browser shows. Depending on your camera and operating system, you may find them in the system camera settings, in a control panel for the camera, in a companion app from the camera's maker, or in the video settings of your calling software. Not every camera offers every control, and names vary, but these are the ones you are most likely to see.

Exposure and Auto Exposure

Exposure controls how long the sensor collects light for each frame. In auto mode the camera adjusts it continuously. Switching to manual lets you lock the brightness so it stops pumping lighter and darker as you move, which is useful when a bright background keeps confusing the auto mode. The trade-off is that a long manual exposure in a dim room can lower the frame rate and blur motion, so add light first and only then fine-tune exposure. Some cameras also offer an exposure compensation or brightness slider that nudges the auto exposure up or down without fully locking it.

Gain or ISO

Gain amplifies the sensor signal to brighten the image, at the cost of visible noise. If your camera lets you limit gain, keeping it low and adding light gives a cleaner picture. If the Noise reading on this page stays poor, high gain in a dim room is the usual cause.

White Balance

White balance tells the camera what color "white" is under your lighting. Auto white balance works well in steady light but can drift when the scene changes, for example when you hold up a colored object or when a cloud passes a window. Setting it manually, often as a slider in kelvin or as presets such as daylight, cloudy or incandescent, keeps skin tones consistent. Choose the setting that makes a white sheet of paper look neutral in the preview.

Focus and Autofocus

Fixed-focus webcams are set at the factory for a typical sitting distance, while autofocus models adjust continuously. Autofocus sometimes "hunts" back and forth, especially in dim light or when you hold something up to the camera. If your camera allows manual focus, set it while sitting in your normal position and lock it. If you want to show objects close to the lens, such as documents or products, leave autofocus on and give it more light so it can lock faster.

Anti-Flicker or Powerline Frequency

Many indoor lights flicker at a rate tied to the mains electricity frequency, which is 50 Hz in some regions and 60 Hz in others. If the camera's exposure timing does not match, you may see horizontal bands rolling slowly through the image or a gentle pulsing in brightness. An anti-flicker or powerline frequency setting, usually offered as 50 Hz, 60 Hz or auto, fixes this; choose the value that matches your local mains supply. Some inexpensive LED bulbs and dimmers flicker on their own regardless, and swapping the bulb is then the real fix.

Backlight Compensation, Low-Light Mode and Other Processing

Backlight compensation brightens the middle of the frame when the edges are bright. It helps when you cannot move away from a window, but it tends to blow out the background. Low-light modes may raise gain and slow the frame rate, which brightens the image but can make movement look smeary and the FPS reading lower.

A useful habit is to change one setting at a time and watch the FPS, Sharpness, Noise and Lighting readings here as you do it. Because the page reads the camera stream directly, any change made in the camera's own control software shows up in the preview and the readings. Note that some systems only let one application use a camera at a time, so you may need to make the change in the companion app, close it, and then return to this page to check the result.

One important distinction: the Image Filters on this page (brightness, contrast, saturation, hue and the presets such as Warm, Cool or Vivid) are visual adjustments to this page's preview. They change how the preview looks here and are included in screenshots you take from the main preview, but they do not change the camera itself, the readings the page calculates, or what other people see in a call. To make a lasting change, use the camera's own settings or the controls in your calling app.

Laptop Camera, External Webcam or Phone as a Webcam

If lighting and settings are as good as you can make them and the picture still falls short, the camera itself may be the limit. There are three common options, each with typical strengths and weaknesses. Individual models vary widely, so treat this as general guidance rather than a rule.

Built-In Laptop Cameras

Laptop cameras have to fit into a very thin screen bezel, which limits the size of the lens and sensor. Many are fine in good light and struggle in dim rooms. Their biggest practical weakness is position: the camera tilts with the screen, so finding a good eye-level angle usually means raising the whole laptop.

External USB Webcams

A separate webcam lets you put the lens where you want it, usually on top of a monitor at eye level. Better models often handle mixed lighting more gracefully and may offer more manual controls. When choosing one, look at what matters for your use: a resolution of at least 720p for calls, a steady 30 frames per second in your actual lighting rather than a headline figure, a field of view that suits your room (wide angles show more background), and a mount that fits your screen. Plug it directly into the computer rather than through a crowded hub if you see low or unstable FPS here. Once connected, run the resolution test on this page to see which resolutions it actually delivers to your browser.

Using a Phone as a Webcam

Modern phones often have better cameras than webcams, and several operating systems and apps let you use a phone as a camera for your computer, connected by cable or wirelessly. The main camera on the back usually gives the best image, so you will need a stand that holds the phone with that camera facing you. Points to keep in mind:

• A wired connection is generally more stable than a wireless one; wireless links can add delay or drop frames when the network is busy.
• Phones can get warm during long sessions, and some lower their frame rate or brightness when they heat up. Keep the phone out of direct sun and, if possible, on a charger.
• Notifications and incoming calls may interrupt the video; turn on a do-not-disturb mode first.
• Once the phone appears as a camera on your computer, this page can test it like any other webcam, so you can check the FPS, resolution and lighting readings before your call.

Why Video Calls Can Look Worse Than the Preview Here

A common surprise is that the camera looks sharp and smooth on this page, yet people on a call say you look blurry, blocky or choppy. Both observations can be true at the same time, because the local preview and a call show you two different things.

The preview on this page plays the camera stream directly on your own computer. Nothing is compressed or sent anywhere for the preview, so you are seeing close to what the camera actually delivers. A video call has several extra steps between your camera and the other person's screen:

1. Encoding: your calling app compresses each frame heavily so it can be sent over the internet. Compression discards fine detail and works by describing changes between frames, so it handles still scenes better than motion or busy, detailed backgrounds.
2. Bandwidth adaptation: the app constantly estimates how much data your connection can carry and adjusts. When your upload speed drops, it lowers the bitrate first, then often the resolution and frame rate. That is why the picture can suddenly turn soft or stutter mid-call even though the camera has not changed.
3. Platform limits: many calling services cap the resolution they send, especially in group calls or when many people are on screen, and some depend on the plan or settings you use.
4. The viewer's side: the other person's connection, device and window size also limit what they see. A small thumbnail tile in a gallery view will never show the detail of a full-screen view.
5. Your computer's workload: encoding video takes processing power. If your computer is busy with many browser tabs, screen sharing or background effects, the calling app may reduce quality to keep up.

The Bitrate figure in this tool's stream details is an estimate calculated from your camera's resolution and frame rate, giving a rough idea of how much data that video might need once compressed. It is not a measurement of your network or of any call. Likewise, the Latency readings and latency test measure the delay inside your own computer between a frame being captured and being shown, not the delay across the internet.

Telling a Camera Problem From a Network Problem

Because the preview here does not involve the network, it works as a useful reference point. Compare what you see here with what others report on a call:

If the problem points to the network, these steps usually help most:

• Use a wired Ethernet connection if possible, or sit closer to your Wi-Fi router with fewer walls in between.
• Pause large uploads and downloads on your network during calls, including cloud backups and file syncing.
• Close other video streams and unnecessary browser tabs.
• Turn off background effects in your calling app if your computer is struggling; they need extra processing.

Using Screenshots, Recordings and AI Detection to Fine-Tune Your Setup

Watching yourself live is useful, but it is hard to judge your own image while you are also sitting in it. Captured stills and short recordings let you step back, compare options side by side and see yourself the way others will. This part explains exactly what this page's capture and AI features produce, where the files end up, and how to use them as a practical testing aid.

What the Screenshot Button Produces

The Screenshot button under the main preview takes a single frame from the live camera and saves it straight away as a JPEG image. A few details are worth knowing:

• Resolution: the image is captured at the camera's full delivered resolution, not at the size of the preview on your screen. A 1920×1080 stream gives a 1920×1080 photo, which makes screenshots good for checking fine detail and focus.
• File name and format: files are named camera- followed by a number based on the time of capture, ending in .jpg, so repeated captures do not overwrite each other and sort in the order you took them.
• Mirror and filters: if you have turned on Mirror View or applied image filters, the screenshot includes them, so it matches what you see in the preview.
• Overlays: the grid lines and AI detection boxes are drawn on a separate layer and are not included in the screenshot.
• Thumbnail: a small preview of the latest shot appears on the page. The Delete Screenshot button removes that thumbnail from the page; it does not delete the file that has already been saved to your device.

The Photo Booth tool works slightly differently. It applies its own filter (None, Gray, Sepia, Blur, Vivid or Dark) and adds each capture to an on-page gallery that holds your most recent photos. Booth photos are not saved automatically; click a thumbnail to download it as a JPEG. Clear Gallery removes the photos from the page.

What the Recording Features Produce

The page offers two ways to record video with sound, and both use your browser's built-in recorder.

The Record Button Under the Main Preview

Record Audio + Video captures the live camera together with your microphone. If the microphone is not already in use by the page, your browser will ask for microphone permission; if you refuse, recording does not start, and the page shows a message explaining why. A timer shows how long you have been recording. When you click Stop Rec, the clip replaces the live view in the same window and plays back in a loop with standard video controls so you can review it immediately.

During playback you have three choices: Download saves the clip to your device, Remove deletes it from the page and returns to the live camera, and Back to Live (or double-clicking the video) also returns to the live camera. The clip only exists in the page's memory until you download it, so download it before you return to the live view if you want to keep it, because going back to live or removing it releases the recording.

The file format depends on your browser. Browsers that support WebM recording save a .webm file, using the VP9 or VP8 video codec where available. Safari, which records in MP4 instead, saves an .mp4 file. Files are named recording- followed by a time-based number. If your browser cannot record video at all, the page tells you rather than failing silently.

The Recording Studio

The Recording Studio in the Tools section is a separate recorder that uses the same camera stream. It shows a timer, keeps a preview player of your last clip below the controls, and has its own Download and Remove buttons. It records in WebM only, so it is intended for browsers with WebM support; Safari users should use the main Record button instead. Studio files are named studio- followed by a time-based number and end in .webm.

What Ends Up in a Recording

• The recording contains the camera image as the camera delivers it. Preview-only effects such as image filters, the grid overlay and AI detection boxes are not burned into the video.
• Mirror View flips only the on-screen preview, so the recorded file shows the camera's normal, unflipped orientation. Text you hold up will read correctly in the file even if it looks reversed in a mirrored preview.
• The microphone is requested with the browser's echo cancellation, noise suppression and automatic gain options turned on, which is similar to how many calling apps capture audio. That makes recordings a reasonable preview of how you will sound on a call, though each calling app processes audio in its own way.
• WebM files play in most modern browsers and many media players. If an editor or player does not accept WebM, a free video converter can change the format.

Where Your Files Go and Who Can See Them

Screenshots, booth photos and recordings are all created inside your browser from the camera stream on your device. When you save one, the browser hands the file to its normal download process, so it lands wherever your downloads usually go, typically your Downloads folder unless you have chosen another location or set the browser to ask each time. The page does not upload captures anywhere as part of saving them.

Once a file is on your device, it is an ordinary file and your usual privacy habits apply:

• Captures can show more than you expect: the room behind you, documents on a wall, other people passing. Review a file before you share it.
• If you are on a shared or work computer, remember that your downloads folder may be synced to cloud storage or visible to other users. Delete test captures you no longer need.
• Unsaved recordings and on-page thumbnails are held only in the page's memory and disappear when you remove them, return to live (for main recordings), or close the tab.

Comparing Lighting Setups With Captures

Judging lighting by eye in real time is surprisingly unreliable because your eyes adjust to whatever you look at. Fixed captures solve this. A simple method:

1. Sit in your usual position and keep the camera fixed in place throughout, so the only thing that changes between captures is the light.
2. Take a screenshot with your current lighting. Note the Lighting, BG Light, Color Temp and Noise readings at the same moment.
3. Change one thing only: turn the desk, move a lamp, close a blind, or add a white card as fill. Wait a few seconds for the camera's auto exposure and white balance to settle.
4. Take another screenshot and note the readings again.
5. Repeat for each option you want to try, then open the images side by side on your computer and compare skin tone, shadows under the eyes, detail in your hair and how bright the background is.

Because screenshots are saved at full resolution, zooming in shows noise and focus differences that are hard to see in a small preview. Keep filters and Mirror View off during comparison shots so you are judging the camera's own image.

For things that change over time, such as auto exposure "breathing" when you move, flicker from lights, or white balance drifting, record a short clip of ten to twenty seconds instead. Move naturally, lean toward and away from the camera, and raise a hand. Play the clip back and watch for brightness pumping, color shifts, rolling bands or blur on movement. Recording also lets you check audio and video together, which a screenshot cannot.

What the Optional AI Detection Does

The AI Detection button opens a panel where you switch on detection as a whole and then choose individual features: Face Detection (a box around each face with a confidence score), Face Mesh (detailed landmarks with eye and mouth tracking), Hand Detection (a 21-point hand skeleton), Body Pose (a 33-point skeleton) and BG Segmentation (a colored mask that separates you from the background). When it is active, the page can also show live labels for blinks, eye openness, smile, head tilt, an estimated expression label and a simple attention indicator.

All detection runs on your own device, inside the browser. The face detection, landmark and expression models are built into the page itself, and the additional hand, pose, mesh and segmentation components are downloaded to your browser from public code libraries when you turn those features on. In both cases the camera frames are analyzed locally and are not sent to a server for detection. Because the work happens on your computer, it uses extra processing power; the page shows a note that a slight FPS drop is normal while AI is active, and that this does not reflect your camera's real performance. Turn detection off before you read the FPS figures if you want a clean measurement.

Detection also feeds into other parts of the page. The Video Call Readiness test can use it for a more dependable Face Framing check than your browser's built-in face detector, and the Quality Analyzer adds a face visibility check when AI is on. If more than one face is detected, the page warns that results may be inaccurate, since the framing and centering checks assume a single person.

How Lighting, Angle and Distance Affect Detection

Face and hand detection models learn from images of faces and hands, and they work best when the picture looks like a clear, well-lit, front-facing view. The same lighting improvements that help your viewers also help detection. Common reasons a face is missed or detection flickers on and off:

• Low light or heavy noise: features such as eyes and mouth become hard to distinguish from grain.
• Strong backlight: a face in silhouette has little visible detail for the model to work with.
• Harsh side light: half the face in deep shadow can confuse landmark placement, which affects blink and smile readings.
• Steep angles: turning your head well to the side, looking down at a laptop from above, or tilting sharply reduces the visible features.
• Distance: to keep detection fast, the page analyzes a reduced-size copy of each frame. A face that is small in the frame, such as when you sit far back from a wide-angle camera, ends up only a few pixels across in that copy and is more likely to be missed.
• Partial covering: hands in front of the face, a microphone, large glasses with reflections, hats or hair over the eyes can all interfere.

This makes detection a handy, if indirect, check of your setup. If the face box stays steady while you move naturally, your lighting and framing are probably reasonable. If it drops out whenever you lean back or turn slightly, that is a hint to add light to your face or move closer to the camera.

Limits of the AI Readings

The detection features are designed as live visual aids, not measurements you should rely on for decisions about yourself or anyone else. Keep these limits in mind:

• Expression labels (such as neutral, happy or surprised) are estimates of facial appearance from a single frame. They do not know how anyone feels, and a neutral resting face, glasses or lighting can easily change the label.
• Blink and eye-openness readings are based on the shape of the eye outline in landmarks. Squinting, glasses, side lighting or looking down can be counted as blinks, and quick blinks can be missed.
• Smile detection compares the shape of your mouth with your recent baseline, so it can react to talking or laughing as well as smiling.
• The attention indicator is a simple score built from whether a face is found near the center with eyes open. It cannot tell whether someone is actually paying attention.
• Left and right hands, pose and segmentation can swap, jitter or bleed into the background, particularly with busy backgrounds, low light or fast movement.
• Results vary between devices and browsers, and detection performance depends on your computer's processing power.

Treat the AI overlay as a tool for checking framing and lighting, or for exploring what in-browser computer vision can do. It is not suitable for monitoring other people, judging mood or making any assessment about a person.

Preparing for Specific Situations

Different situations call for slightly different priorities. The checklists below assume your camera already works in the test above; they focus on how to use the page's readings and capture tools to get ready.

Job Interviews

An interview is where small details carry the most weight, and you usually have time to prepare. A day or two before:

• Set up in the room and at the time of day you will use, so the light matches. Run the Video Call Readiness test and work through its suggestions until lighting, framing and distance are no longer flagged.
• Raise the camera to eye level and place the interviewer's video window directly under the lens.
• Choose a plain, tidy background with nothing personal or confidential readable. Check it in a full-resolution screenshot.
• Record a short clip answering a practice question. Watch it back for lighting, posture, how often you look away from the lens, and how your voice sounds. If the audio needs work, the microphone test lets you check levels on their own.
• Check clothing on camera; fine stripes and small checks can shimmer on video, and very bright white or very dark clothing can push the camera's exposure.
• On the day, open the test page a few minutes early to confirm the camera starts and the FPS is steady, then close it before joining so the calling app has the camera to itself.

Online Classes

Students and teachers often spend hours on camera, sometimes in shared spaces and at different times of day.

• Plan for both daylight and evening sessions. A desk lamp with a matching bulb color gives consistency when the window light fades.
• Keep the camera at a distance that shows your face and hands if you will be writing or demonstrating.
• For long sessions, a stable mount matters more than resolution; the Stability reading will flag a camera that is shaking with a wobbly desk or typing.

Streaming

Streams are watched for longer and often at full screen, so image quality and consistency are more noticeable.

• Use a full key, fill and back light arrangement if you can. A back light helps separate you from a dark room, which matters if your streaming software uses background removal.
• Lock exposure and white balance in your camera's settings so the image does not shift when bright game scenes reflect onto your face from the monitor.
• Check the anti-flicker setting if you have LED strips or RGB lighting in the shot; some cause rolling bands.
• Use the Grid Overlay to place yourself where your stream layout leaves room for the camera, such as a corner of the screen, and check headroom with that crop in mind.
• Record a short clip and watch it full screen to see how your camera looks at stream size.

Remote Work Meetings

Daily meetings reward a setup that is good enough every time without fiddling.

• Fix the camera and lights in place so the setup is the same every day, and avoid seating positions where a window ends up behind you as the sun moves.
• Keep a quick routine: open the test, confirm the picture and the FPS, glance at the background for anything that should not be on view, then join.
• If colleagues say your video is poor but it looks good here, use the comparison table above to check whether the network is the cause before buying new hardware.
• When you share your screen, your calling app may lower your camera quality to prioritize the shared content; that is normal behavior, not a camera fault.

Recording Tutorials and Demonstrations

Tutorials are viewed repeatedly and often paused, so clarity and consistency over the whole recording matter.

• Light both your face and anything you demonstrate. Objects held close to the camera need more light so autofocus can lock on them.
• Lock focus at your normal position if you mostly talk to camera, or leave autofocus on if you frequently show items close to the lens, and test both with a short clip.
• Keep the whole recording in the same lighting. If you record over several sessions, take a reference screenshot of your setup on the first day and compare against it each time you start.
• Record a test clip with sound, then play it back to check audio sync, background noise and whether your hands or props leave the frame.

🔐 Webcam Privacy and Security: A Practical Guide

A webcam is one of the more personal devices attached to your computer, so it is worth knowing how access to it is controlled. The good news is that modern browsers and operating systems put several layers between a website and your camera, and most of the protection comes from a few simple habits. This guide explains how those layers work, what the warning signs of unwanted access look like, and how to keep control without becoming anxious about every blinking light.

How Browser Camera Permissions Work

A web page cannot simply switch on your camera. When a site wants video, it has to ask the browser, and the browser asks you. That permission is tied to the site's origin, meaning its exact address (for example, the domain and the secure https scheme). Allowing camera access on one site does not give access to any other site, including other pages that look similar or claim to be related.

A few properties of the permission system are useful to understand:

• It is per site. Each website gets its own decision. Your choice for a video-call service has no effect on a shopping site, and vice versa.
• It is separate for camera and microphone. A site can have camera access without microphone access. On this page, for example, the live preview only asks for video. The microphone prompt appears only when you press Record Audio + Video, because recordings include sound. If you decline the microphone, the recorder does not start, but the camera preview keeps working.
• It can be temporary or remembered. Depending on your browser, you may see options such as allowing access once, allowing it for the current visit, or remembering the choice. A one-time choice is a sensible default for sites you rarely use.
• It is revocable at any moment. You can change the decision later from the browser's site settings without affecting anything else.
• It requires a secure connection. Browsers only expose the camera to pages loaded over HTTPS (or to local development pages), which is why camera tools on plain HTTP pages fail.
• The page must be active. Browsers generally only show the permission prompt in response to a page that is open and visible, and many will not let a background tab start the camera on its own.

Even after you allow access, a well-behaved site should only open the camera when you ask it to. This camera test lists your cameras when the page loads, but it does not start the stream until you click ▶ Test My Camera. If you have never granted permission before, camera names may appear blank or generic in the dropdown until you allow access, because browsers hide device labels from sites that have not been granted permission.

Indicator Lights and On-Screen Camera Indicators

There are usually several independent ways to tell whether your camera is in use, and it helps to know all of them.

The Hardware Light

Most webcams and laptop cameras have a small light next to the lens that turns on while the sensor is streaming. On many devices this light is wired so that it comes on whenever the camera has power, but that is not true of every model, and on some cameras the light is controlled by firmware. Treat the light as a strong hint rather than an absolute guarantee. When you click ■ Stop on this page, the tool releases every camera track it opened, and the light should go out within a moment. If it stays on after you stop the test and close the tab, another program is probably using the camera.

Browser Indicators

Desktop browsers typically show a camera icon in the address bar, and often a recording dot or icon on the tab itself, whenever a page is capturing video. Clicking the address-bar icon usually shows which device is in use and lets you block access for that site. If you have many tabs open, scan them for that icon before assuming a call or test is fully closed.

Operating System Indicators

Recent versions of the major desktop and mobile operating systems also show their own indicator when any app uses the camera, such as a small colored dot or a camera icon in the menu bar, status bar or system tray. Many systems also keep a list of which apps accessed the camera recently, found in the privacy section of the system settings. That list is one of the most reliable ways to answer the question "what has been using my camera?", because it covers desktop apps as well as websites.

Signs of Unwanted Camera Access (and Innocent Explanations)

Most "my webcam light turned on by itself" moments have ordinary causes. Before worrying, check the common explanations:

• A video-call app that was minimized rather than closed, or one that opens a camera preview when it starts with your computer.
• Facial sign-in features, which briefly activate an infrared or regular camera on the login screen or when waking the computer.
• Vendor camera utilities or virtual-camera software that holds the device open in the background.
• A browser tab you forgot about, such as a meeting page or another camera test.
• Driver or system updates that briefly initialize hardware.

Signs that deserve a closer look are different in kind: the camera light turning on repeatedly when no app you recognize is running, an unfamiliar program appearing in your operating system's camera-access history, new remote-access or "support" software you did not install, security settings that have been switched off without your knowledge, or your camera showing as "in use by another application" when nothing obvious is open. None of these proves anything on its own, but together they justify a security scan and a review of installed programs.

A quick check with this page can help narrow things down. If you click ▶ Test My Camera and get a black screen or an error saying the device is busy while no app appears to be open, something else currently holds the camera. Use your operating system's camera-access history or task manager to find out what it is.

Malware, Remote-Access Scams and Extortion Emails

Unwanted camera access almost always comes from software running on your own computer, not from a website you visited. In general terms, the risks fall into three groups.

Remote-Access Malware

Some malicious programs give an attacker remote control of an infected machine, which can include the camera and microphone. They usually arrive the same way other malware does: an email attachment, a pirated program or game crack, a fake software update, or a download from an untrusted site. The defenses are the familiar ones: keep your operating system and security software updated, install programs only from sources you trust, and be suspicious of attachments you did not expect.

Fake Tech-Support Calls and Pop-Ups

A common scam involves a pop-up, phone call or message claiming your computer is infected or your account is compromised, followed by a request to install remote-desktop software so a "technician" can fix it. Once installed with your permission, that software can see your screen and use your devices. Legitimate companies do not cold-call you to fix your computer. If you have already installed something at a stranger's request, disconnect from the internet, uninstall it, run a security scan and change important passwords from a different, trusted device.

"I Recorded You" Extortion Emails

Another widespread scam is an email claiming the sender hacked your webcam and recorded you, demanding payment to keep the footage private. These messages are typically sent in bulk and are usually bluffs; to sound convincing, they sometimes quote an old password taken from a past data breach on some other website. Do not reply and do not pay. If the email includes a password you still use anywhere, change it on every account that uses it and turn on two-factor authentication. You can report the message to your email provider and, if you wish, to your country's fraud-reporting service.

Practical Habits That Keep You in Control

Cover or Shutter the Lens When Not in Use

A physical cover is the simplest guarantee that nothing useful can be seen. Many external webcams and some laptops include a built-in privacy shutter; for others, a slide-on cover works. On laptops, check that a stick-on cover is thin enough for the lid to close fully, since a thick cover can press on the screen. A cover does not block the microphone, so it complements rather than replaces good permission habits. You can confirm a shutter works by starting this test and closing it: the preview should go black, and the lighting reading should drop sharply.

Review and Revoke Site Permissions

Every few months, open your browser's site settings and look at the list of sites allowed to use your camera and microphone. Remove any you no longer use. For this page, you can revoke access the same way after testing; the tool will simply ask again next time. The process is usually:

1. Click the icon at the left of the address bar (often a padlock or settings symbol) while on the site.
2. Open the site settings or permissions entry.
3. Set Camera (and Microphone) to Ask or Block, or reset the permissions for that site.
4. Reload the page so the change takes effect.

Review App Permissions in Your Operating System

Desktop and mobile operating systems have their own camera privacy settings, separate from the browser. They control which apps (including the browser itself) can use the camera at all. Remove access for apps that do not need it. Remember that if you turn off camera access for your browser at the system level, no website will be able to use the camera, including this test, until you turn it back on.

Keep Software Updated

Browser, operating system and camera driver updates regularly include security fixes. Turning on automatic updates for the browser and operating system covers the most important parts. If your webcam has optional vendor software, update it too, or uninstall it if you never use its features; a standard webcam works without it.

Close Calls and Tests Completely

When you finish a call or a test, use the app's leave or stop button and close the tab, rather than just switching away. On this page, ■ Stop releases the camera immediately and clears the screenshot thumbnail and recording preview held in the page's memory.

Using a Webcam Safely on Shared or Public Computers

Library, school, office and hotel-lobby computers need a little extra care, because the next person sits down at the same browser and the same downloads folder.

• Use a private or incognito window. Permissions granted in a private window are generally discarded when the window closes, so the next user does not inherit your camera access.
• Decline "remember this decision". If the browser offers to remember your choice, choose a one-time option instead.
• Delete downloaded files. Snapshots and recordings from this tool are saved to the browser's download location, which on a shared machine may be visible to everyone. Files from this page are named with a prefix such as camera-, booth-, recording- or studio- followed by a long number, which makes them easy to find and remove. Empty the recycle bin or trash afterward if you can.
• Clear the browser's download list. Even after deleting a file, the download history can still show its name.
• Close every window when you leave. Closing the browser ends any camera streams and private sessions.
• Check the camera before you sit in front of it. On an unfamiliar machine, look for an indicator light or an on-screen camera icon before starting, and use the lens cover if one is fitted.

On a work or school computer, also remember that the device belongs to your organization, which may run its own monitoring or management software. Follow its policies on cameras and recordings.

How This Camera Test Handles Your Camera

Based on how the page is built, here is exactly what happens to video when you use it:

• Camera on only when you ask. The page enumerates your devices so the dropdown can list cameras, but the stream starts only after you click ▶ Test My Camera and approve the browser prompt.
• Analysis happens in the page. Frame rate, sharpness, lighting, noise, stability and the optional AI features (face, landmarks, hands, pose and background segmentation) are calculated by scripts running in your browser tab, working on frames drawn to an off-screen canvas. The frames themselves are not sent anywhere.
• Snapshots and recordings stay on your device. Screenshots, photo-booth captures and recordings are created in the browser's memory and then downloaded to your device. They are not uploaded. Whether they ever leave your device is up to you, for example if you choose to attach one to an email or share it in a chat.
• Recording stops if you switch away. If you hide the tab or minimize the browser while the main recorder is running, recording stops automatically.
• Stop really stops. The Stop button ends every camera track the page opened, so the camera light should turn off, and it releases the in-memory snapshot and recording previews.
• What the page does load. Like most websites, the page downloads code and AI model files from public content-delivery networks so the features can run locally, and it uses a basic analytics counter that records events such as a test being completed. That counter receives the event name and the tool name, not images, video, audio or measurement results.
• Share buttons share the link only. The WhatsApp, X, Facebook and Copy Link buttons share the page address, not your snapshots or recordings.

If you want extra reassurance, you can watch your browser's camera indicator while using the page and confirm it disappears when you click Stop, and you can revoke the site's camera permission when you are done.

Privacy Considerations When You Share Snapshots or Recordings

Once a file is on your device, ordinary file-sharing common sense applies. A frame from your webcam often shows more than your face:

• Documents, sticky notes with passwords, or screens visible behind you.
• Family photos, mail with your address, or windows that reveal your location.
• Other people in the room who did not agree to be filmed.
• Reflections in glasses, mirrors and glossy screens.

Before sending a test clip to a support agent or colleague, play it back in the preview first and check the background and audio. If you only need to show a technical problem such as flicker or dropped frames, point the camera at a plain wall or object rather than yourself. Ask permission before recording anyone else, and be aware that rules about recording other people vary by place.

Webcam Privacy Checklist

1. Keep the lens covered or the shutter closed when you are not using the camera.
2. Allow camera access only to sites you trust, and prefer one-time permissions.
3. Review browser site permissions and operating-system app permissions every few months.
4. Keep your browser, operating system and camera software up to date.
5. Learn what your camera light and on-screen indicators look like, and check them when a call ends.
6. Never install remote-access software at the request of someone who contacted you unexpectedly.
7. Ignore webcam extortion emails, and change any password they quote.
8. On shared computers, use a private window and delete downloaded snapshots and recordings.
9. Check the background of any clip or snapshot before sharing it.

🛒 Choosing a Webcam and Setting It Up for Everyone

Webcam spec sheets can be confusing, and the biggest number on the box is rarely what makes the picture look good. This section explains which specifications actually matter, how to check a new camera with the tools on this page, what to look at before the return window closes, and how to place and light a camera so video calls work well for everyone, including people who use captions, lip-read, sign or work from a wheelchair or standing desk.

Start With How You Will Use It

Before comparing specifications, decide what the camera is for. The answer changes which features are worth paying attention to:

• Everyday video calls: Good low-light performance, sensible field of view and reliable plug-and-play matter more than very high resolution, because most call apps compress video heavily anyway.
• Streaming and content creation: Higher frame rates, good autofocus and support for manual control of exposure and white balance become more useful.
• Showing documents or objects: Autofocus that works at close range is important, since fixed-focus cameras often blur anything held close to the lens.
• Group rooms: A wider field of view and better microphones matter more than fine detail on one face.
• Laptop users: Weight, a mount that grips a thin lid securely and a short cable may matter more than any image specification.

Webcam Specs That Actually Matter

Sensor Size and Low-Light Performance

The image sensor turns light into a picture. In general, a larger sensor (and larger individual pixels) can gather more light, which tends to mean less grain and better color indoors. Lens quality and the camera's image processing play a big part too. Sensor size is not always listed, so look for honest descriptions of low-light behavior and, after buying, test the camera in the room where you will actually use it. Most rooms in the evening are far darker than a camera would like, and that is where differences between webcams show most.

One behavior to know about: many webcams automatically lower their frame rate in dim light so each frame can be exposed for longer. The picture gets brighter but motion becomes choppier. If your FPS reading on this page drops when you turn off a lamp and recovers when you turn it back on, that is usually the camera doing this deliberately, not a fault.

Resolution Versus Frame Rate

Resolution describes how many pixels each frame has; frame rate describes how many frames arrive each second. More resolution gives finer detail, while more frames give smoother motion. Cameras often cannot deliver their highest resolution and highest frame rate at the same time, so a camera advertised with both a high resolution and a high frame rate may only reach the higher frame rate at a lower resolution. Read the specification as pairs, for example "1080p at 30 fps" or "720p at 60 fps", rather than as separate numbers.

For most video calls, 1080p or 720p at around 30 frames per second is plenty, because the call app will usually scale the video down to fit the connection. Higher resolutions are more useful for recording, streaming or cropping into the picture.

Field of View

Field of view (FOV) is how wide an area the camera sees, usually given in degrees and usually measured diagonally. A narrower field keeps the frame on you and hides more of the room. A wider field fits several people or a larger space but shows more background and can make faces near the edges look stretched. For a single person at a desk, a moderate field of view is usually the most flattering; for meeting rooms, wider is often better. Some cameras let you choose between fields of view in their own software.

Autofocus Versus Fixed Focus

Fixed-focus cameras are set to keep things sharp at a typical sitting distance and beyond. They are simple and never "hunt", but anything held close to the lens, such as a document or a product, may look soft. Autofocus cameras adjust for distance, which helps if you move around or show objects, though a slow or nervous autofocus can pulse in and out of focus. If you show things on camera often, autofocus is worth having; if you sit still, a good fixed-focus camera is fine.

Built-In Microphones

Many webcams include one or two microphones. They are convenient, but because they sit a meter or so from your mouth, they also pick up room echo, keyboard noise and fans. For regular calls, a headset or a separate microphone near your mouth usually sounds clearer. A webcam microphone is still useful as a backup or for casual calls. You can check one quickly with the microphone test or by making a short recording on this page.

Mounting and Positioning Options

Look at how the camera attaches. A good monitor clip holds firmly on both thin laptop lids and thicker monitors without tipping forward. A standard tripod thread on the base lets you use a small tripod, a desk arm or a monitor-mounted bracket, which is especially helpful when you need the camera at an unusual height. A tilt hinge lets you aim the lens without moving the whole mount, and a swivel lets you turn it toward a whiteboard or another person.

USB Connection and Cable

Most webcams connect over USB, using either a USB-A or USB-C plug. Make sure the plug matches a free port on your computer, or plan for an adapter. Many 1080p webcams work fine on USB 2.0 because they compress video inside the camera before sending it, while some higher-resolution models need a faster port to reach their full specification. Cable length matters if your computer sits under the desk; extension cables and unpowered hubs are a common source of dropouts and low frame rates, so test the camera plugged directly into the computer first.

Privacy Shutter

A built-in sliding shutter or a lens cap is a small feature that is easy to appreciate later. It lets you block the lens without unplugging the camera, and it is more convenient than a separate stick-on cover. Check that it closes completely and does not rattle open.

Operating System Compatibility and UVC

Most webcams follow the USB Video Class (UVC) standard, which lets them work with the generic camera drivers built into the major desktop operating systems. In practice, that means you plug the camera in and it appears in browsers and call apps without installing anything. Vendor software, if offered, is usually optional and adds extras such as zoom, field-of-view presets, auto-framing or saved image settings. If a camera requires its own driver or app just to function, check that it supports your operating system and version before buying, and be aware that it may not work at all on systems the vendor does not support.

Image Controls and HDR

Automatic exposure and white balance are fine for most people, but the option to lock them helps when lighting in your room changes during a call or when the camera keeps adjusting as you move. Features marketed as HDR or wide dynamic range help when you sit in front of a bright window, keeping both your face and the background visible. These controls are usually reached through the camera's own software or your operating system's camera settings rather than through the browser.

Spec Sheet Pitfalls to Watch For

• Megapixels quoted for photos: A camera may advertise a still-photo resolution that is higher than the resolution it can stream as video.
• "Up to" frame rates: The highest frame rate is often only available at a lower resolution or in bright light.
• Interpolated resolution: Some cameras enlarge a lower-resolution image in software, which adds pixels but not detail.
• Unlisted field of view: If the field of view is not stated, sample images or videos are the best guide to how much of the room you will show.
• Features that require an app: Auto-framing, background effects and image presets may only work while the vendor's software is installed and running.

Testing a New Webcam When It Arrives

The first few days with a new camera are the best time to find problems, while it is easy to return. The tools on this page can cover most checks in about fifteen minutes. Work through them in the room and lighting where you normally use the camera.

1. Plug it straight into the computer. Avoid hubs and extension cables for the first test. Once the camera works, you can add them back one at a time to see whether they cause problems.
2. Check that the right camera is detected. Click 🔄 Refresh next to the camera dropdown if needed, select the new camera and start the test. Compare the camera name in the Device Info and Camera Information panels with what you expect, so you know you are not testing the laptop's built-in camera by mistake.
3. Run the Resolution Test. It checks which of the standard sizes (1920 × 1080, 1280 × 720, 640 × 480 and 320 × 240) the camera accepts through the browser. A camera sold as Full HD should pass 1920 × 1080; if it does not, try a different USB port before concluding there is a fault.
4. Watch the live FPS in good and dim light. Note the FPS with the room well lit, then dim the lights. A moderate drop in dim light is normal for many webcams; a consistently low frame rate in bright light is worth investigating.
5. Check focus at different distances. Use the Quality Analyzer or the live Sharpness reading while sitting at your normal distance, then hold a printed page closer to the lens. Autofocus cameras should sharpen after a moment; fixed-focus cameras will usually stay soft up close, which is expected.
6. Compare old and new side by side. If you are replacing a camera, open the Camera Comparison tool, choose the old camera as A and the new one as B, and drag the divider across the image. The per-side FPS, brightness, sharpness and noise readings help you see whether the upgrade is real in your lighting.
7. Measure latency (Chrome or Edge). The Latency Test estimates the delay from capture to display. Run it on both cameras if you are comparing, since a camera that adds noticeable delay will make your lip movements lag behind your voice on calls.
8. Make a short recording. Use the Recording Studio to record ten or twenty seconds of you talking and moving your hands, then play it back. Listen to the built-in microphone, if the camera has one, and look for stutters, color shifts or flicker.
9. Run the Video Call Readiness test. It combines frame rate, resolution, lighting, sharpness, noise, stability and face framing into a single score with suggestions, which is a helpful summary of how the camera performs in your setup.
10. Test the shutter and the light. Close the privacy shutter and confirm the preview goes dark. Then click ■ Stop and confirm the camera light turns off.
11. Try unplugging and replugging. Reconnect the camera while the page is open and check that it reappears in the dropdown, which confirms the connection and driver handle reconnection properly.
12. Test in the apps you actually use. Finally, open your usual video-call app and confirm the new camera is selected there, since each app keeps its own camera choice.

Save a snapshot or short recording from your first test. It gives you a reference for how the camera looked when new, which is useful if you later suspect it has changed.

Warranty and Return Checklist

Do these things before the return window closes, and keep notes as you go:

• Keep the receipt and packaging until you are sure the camera is staying, including the clip, any cap or cover and the cable.
• Note the return deadline and the warranty period. Both are usually in the retailer's terms or the manufacturer's documentation.
• Check for physical defects: scratches or dust inside the lens, a loose hinge, a clip that will not hold, a shutter that will not close, or a cable that only works at certain angles.
• Look for stuck or hot pixels. With the lens covered or in a dark room, bright fixed dots that stay in the same place in the preview can indicate sensor defects. Some faint noise in darkness is normal; a dot that never moves is not.
• Confirm the advertised modes. Make sure the resolution, frame rate and field of view you paid for actually appear on your computer, using the Resolution Test and live readings.
• Test on a second computer if possible. If a camera fails on one machine but works on another, the cause is more likely a port, driver or setting than the camera itself.
• Record evidence of any fault. A short recording or snapshot showing the problem, plus a note of your operating system and browser, makes a support request much easier.
• Check for firmware updates from the manufacturer's official support site before deciding a fault cannot be fixed.
• Register the product if the manufacturer requires registration for warranty service.

Accessibility and Comfort on Video Calls

Camera setup affects more than how you look. Good placement and lighting make it easier for others to read your lips and expressions, follow sign language and feel that you are engaged, and a comfortable setup makes long calls less tiring for you. The suggestions below are practical starting points; adapt them to your own needs and equipment.

Captions

Many video-call platforms offer live captions, and many operating systems and browsers now include their own live-captioning features that work across apps. Captions help people who are deaf or hard of hearing, people listening in a second language, and anyone on a call with poor audio. A few habits make automatic captions more accurate for everyone: use a microphone close to your mouth, reduce background noise, speak one person at a time and say names and technical terms clearly. If a meeting is important, check in advance whether participants need captions or a human captioner, and test that captions are turned on before the meeting begins.

Lighting for Lip-Reading and Expression

People who lip-read, and everyone else to some degree, rely on seeing the mouth and face clearly. Lighting makes the biggest difference:

• Light your face from the front, for example with a lamp or window in front of you and slightly above eye level.
• Avoid sitting with a bright window or lamp behind you, which turns your face into a silhouette.
• Use soft, even light rather than a single harsh spotlight that casts deep shadows across your mouth.
• Keep your mouth uncovered and visible: avoid resting your hand on your chin or holding a microphone in front of your lips.
• Aim for a steady frame rate, since choppy video makes lip movements harder to follow. Bright light usually helps the camera keep a higher frame rate.

You can use the Lighting reading on this page as a quick check: adjust lamps until it reports good lighting with your face clearly visible, and use the Video Call Readiness suggestions for further hints.

Framing for Sign Language

Signers need more of the frame than a typical head-and-shoulders shot. Frame from just above your head to around your waist, with enough room on both sides for your hands. A plain background that contrasts with your skin and clothing makes hand shapes easier to see, and good lighting with a steady frame rate keeps fast movements readable. With AI Detection turned on, the Hand Detection overlay and the Hands Detected reading on this page can confirm that both hands stay inside the frame while you sign.

Camera Placement for Wheelchair Users

Laptops and monitor-mounted webcams are usually set up for someone sitting in a standard desk chair, which may put the camera too high or too low for a wheelchair user. Some options:

• Mount the camera on a small tripod, desk clamp or flexible arm so it can sit at your eye level regardless of the desk height.
• If you use a laptop on a lap tray, raise the laptop on a stand or use an external camera, so you are not filmed from below.
• Choose a camera with a tilt hinge or a mount that adjusts with one hand, and test that you can reach the privacy shutter easily.
• If your chair height changes, for example with a tilt or elevating seat, re-check framing after adjusting.
• Keep cables tidy and clear of wheels and footrests.

Standing and Sit-Stand Desks

If the camera clips to a monitor that sits on the desk, it moves up and down with the desk, which keeps framing roughly consistent. If the monitor is on a separate arm or the camera is on a fixed stand, switching between sitting and standing can leave you looking at a camera pointed at your chest or above your head. After changing desk height, glance at the preview and adjust the tilt. The Grid Overlay on this page helps you get your eyes near the upper third of the frame quickly, and the Face Position reading tells you whether you are centered.

Eye Contact Tips

On a video call, looking at someone's face on screen means you are not looking into the lens, so they may feel you are glancing down. A few ways to close that gap:

• Place the camera as close as possible to where faces appear on screen, usually centered on top of the monitor you use for calls.
• Move the call window or the active speaker's video to the top center of the screen, just below the camera.
• Look into the lens when you are making an important point or asking a question, and look at people's faces when they are speaking. You do not need to stare at the lens all the time.
• Sit far enough back that small eye movements to read notes are less noticeable.
• If you use two monitors, put the camera on the one where the call window lives.

Eye contact customs vary between people and cultures, and some people find sustained eye contact uncomfortable. A camera at eye level and a natural, relaxed gaze matter more than perfect eye contact.

Comfort on Long Calls

• Mirrored self-view is normal. Most call apps, and this page's preview, show you a mirror image, so text on a shirt or a sign looks reversed to you but usually appears correctly to others. The Mirror Tool lets you flip the preview to see yourself as others typically do.
• Hide your self-view once your framing is set, if seeing yourself all the time is distracting; most call apps offer this.
• Reduce glare on glasses by raising or angling the light source to the side, and by lowering screen brightness slightly.
• Watch for flicker. Some lights flicker at a rate the camera picks up as bands or pulsing, which can be unpleasant to watch. If you see banding, try a different light or set the camera's anti-flicker option to match your local mains frequency in its software.
• Take breaks. Stand up, look at something far away and rest your eyes between calls. If you feel discomfort, stop and rest.

📖 Webcam and Video Glossary

The terms below appear on webcam boxes, in camera software and in the readings on this page. Definitions are kept short and general; individual cameras and apps may use some terms slightly differently.

Image and Sensor Terms

Aperture

The opening in the lens that lets light through, written as an f-number such as f/2.0. A smaller f-number means a wider opening and more light reaching the sensor. Most webcams have a fixed aperture.

Aspect ratio

The shape of the frame, written as width to height. 16:9 is the common widescreen shape (for example 1920 × 1080); 4:3 is a squarer shape used by older and some lower-resolution modes (for example 640 × 480).

Autofocus

A system that adjusts the lens automatically to keep the subject sharp at different distances. Cameras without it are called fixed focus.

Backlighting

Light coming from behind the subject, such as a window behind you. It tends to make faces dark because the camera exposes for the bright background.

Color temperature

A measure of whether light looks warm (orange) or cool (blue), expressed in kelvin (K). Household lamps are often warm, daylight is cooler. Mixing the two can give a face uneven color.

Contrast

The difference between the darkest and brightest parts of an image. Low contrast looks flat and washed out; very high contrast loses detail in shadows and highlights.

Dynamic range

The range from dark to bright that a camera can capture in one frame while keeping detail in both. Scenes with a bright window and a dim face test dynamic range.

Exposure

How much light is captured for each frame, controlled by exposure time and gain. Underexposed images look dark; overexposed images look washed out with lost highlights.

Field of view (FOV)

How wide an area the camera can see, usually stated in degrees and often measured diagonally. The FOV Class shown on this page is an estimate based on the stream's megapixels and facing mode, not a measurement of the lens.

Fixed focus

A lens set at one focus distance that keeps most of a typical scene acceptably sharp without moving parts. Very close objects may look blurry.

Gain (ISO)

Electronic amplification of the sensor signal to brighten a dark image. Higher gain makes the picture brighter but also increases visible noise. ISO is the photography term for a similar sensitivity setting.

HDR (high dynamic range)

Techniques that extend the range of brightness a camera can show, for example by combining information from different exposures, so that both bright and dark areas keep detail. Webcam makers sometimes call this WDR (wide dynamic range).

Hot pixel

A sensor pixel that reads brighter than it should, showing as a fixed bright dot, most visible in dark images. A stuck or dead pixel stays at one color or stays dark.

Megapixel (MP)

One million pixels. A 1920 × 1080 frame contains about 2.1 megapixels. More megapixels allow more detail but do not by themselves guarantee a better image.

Noise

Random grain or speckles in an image, most visible in dim light or when gain is high. The Noise reading on this page estimates how much of this grain is present.

Rolling shutter

A way of reading a sensor line by line rather than all at once. Most webcam sensors work this way, which can make fast movement look skewed or wobbly and can cause banding under flickering lights.

Sensor

The light-sensitive chip that captures the image. Its size and design strongly influence low-light performance and noise.

Sharpness

How clearly fine edges and details appear. It depends on focus, lens quality, motion and processing. The Sharpness reading on this page is based on edge detail in the live frames.

White balance

Adjustment that makes white objects look white under different lighting, correcting warm or cool color casts. It can be automatic or set manually in camera software.

Video, Format and Performance Terms

1080p, 720p and 480p

Shorthand for video heights of 1080, 720 and 480 pixel rows. 1080p usually means 1920 × 1080 (Full HD), 720p means 1280 × 720 (HD), and 480p commonly refers to 640 × 480 on webcams. The "p" stands for progressive scan.

4K

A resolution around 4,000 pixels wide, most commonly 3840 × 2160 for video. It offers more detail for cropping and recording but needs more bandwidth and processing.

Bitrate

The amount of data used per second of video, usually in kilobits or megabits per second. Higher bitrate generally means fewer compression artifacts. The bitrate shown on this page is an estimate, not a measurement of a file.

Codec

The method used to compress and decompress video (or audio). Common video codecs include H.264, VP8, VP9 and AV1.

Compression artifacts

Visible flaws caused by heavy compression, such as blockiness, smearing on moving objects or fuzzy edges around text.

Container

The file format that holds video and audio streams together, such as WebM or MP4. A container is different from the codec used inside it.

Frame rate (FPS)

The number of frames shown or captured each second. Around 30 fps is common for calls; 60 fps gives smoother motion.

Frame drop

A frame that was expected but not delivered or displayed, which shows up as a stutter and as a dip in the measured FPS.

H.264

A very widely supported video codec, also called AVC. Some webcams can compress video to H.264 inside the camera, and many apps and devices can play it.

Jitter

Variation in timing from one frame to the next. Low jitter means frames arrive at an even pace; high jitter feels uneven even when the average frame rate looks fine.

Latency

The delay between something happening in front of the camera and it appearing on screen. The Latency Test on this page measures the capture-to-display delay in the browser, not the extra delay a call adds over the network.

MJPEG

Motion JPEG, a format that sends each frame as a separate compressed JPEG image. Many USB webcams use it to fit higher resolutions through a USB 2.0 connection.

Resolution

The number of pixels in each frame, written as width × height, such as 1280 × 720.

Uncompressed (YUV) video

Video sent from the camera without compression, often in a YUV pixel format such as YUY2. It avoids compression artifacts but needs much more USB bandwidth, so it is usually limited to lower resolutions or frame rates on USB 2.0.

WebM

An open video container format that browsers commonly use for recordings, typically holding VP8 or VP9 video. Recordings from this page are saved as WebM, or MP4 in browsers such as Safari that record in that format.

Connection, Hardware and Software Terms

Anti-flicker

A camera setting that matches exposure to the mains electricity frequency (50 Hz or 60 Hz, depending on the region) to reduce banding from artificial lights.

Auto-framing

A feature, usually in camera software or the camera itself, that digitally zooms and pans to keep a person centered in the frame.

Digital zoom

Enlarging part of the image by cropping it, rather than by moving lens elements. It reduces detail, unlike optical zoom.

Driver

Software that lets the operating system communicate with a device. Most webcams use the generic driver built into the operating system.

Facing mode

Whether a camera faces the user ("user", like a front or laptop camera) or faces away ("environment", like a phone's rear camera). This page shows it when the browser reports it.

getUserMedia

The browser function that asks for permission and opens the camera or microphone for a web page.

Privacy shutter

A physical slide or cap that covers the lens to block the camera's view when not in use.

Tripod thread

A threaded socket on the base of a camera or mount that attaches to standard tripods and arms. Most small cameras use the common quarter-inch size.

USB bandwidth

The amount of data a USB connection can carry. Sharing a port or hub with other busy devices can leave too little for a camera, leading to lower frame rates or failed starts.

UVC (USB Video Class)

A standard that lets USB cameras work with the operating system's built-in drivers, so they work without installing vendor software on supported systems.

Virtual camera

Software that appears to apps as a camera but supplies processed or generated video, such as effects or a screen capture. Virtual cameras may show up in this page's camera list.

WebRTC

A set of browser technologies for real-time audio, video and data. Camera access in browsers, including on this page, is part of this family of features.