The MediaDevices.getUserMedia() method prompts the user for permission to use a media input which produces a MediaStream with tracks containing the requested types of media.
That stream can include, for example, a video track (produced by either a hardware or virtual video source such as a camera, video recording device, screen sharing service, and so forth), an audio track (similarly, produced by a physical or virtual audio source like a microphone, A/D converter, or the like), and possibly other track types.
It returns a Promise that resolves to a MediaStream object. If the user denies permission, or matching media is not available, then the promise is rejected with NotAllowedError or NotFoundError DOMException respectively.
Note: It's possible for the returned promise to neither resolve nor reject, as the user is not required to make a choice at all and may ignore the request.
getUserMedia(constraints)
A Promise whose fulfillment handler receives a MediaStream object when the requested media has successfully been obtained.
As an API that may involve significant privacy concerns, getUserMedia()'s specification lays out a wide array of privacy and security requirements that browsers are obligated to meet.
getUserMedia() is a powerful feature that can only be used in secure contexts; in insecure contexts, navigator.mediaDevices is undefined, preventing access to getUserMedia(). A secure context is, in short, a page loaded using HTTPS or the file:/// URL scheme, or a page loaded from localhost.
In addition, user permission is always required to access the user's audio and video inputs. Only a window's top-level document context for a valid origin can even request permission to use getUserMedia(), unless the top-level context expressly grants permission for a given <iframe> to do so using Permissions Policy. Otherwise, the user will never even be asked for permission to use the input devices.
For additional details on these requirements and rules, how they are reflected in the context in which your code is running, and about how browsers manage user privacy and security issues, read on.
As an API that may involve significant privacy concerns, getUserMedia() is held by the specification to very specific requirements for user notification and permission management. First, getUserMedia() must always get user permission before opening any media gathering input such as a webcam or microphone. Browsers may offer a once-per-domain permission feature, but they must ask at least the first time, and the user must specifically grant ongoing permission if they choose to do so.
Of equal importance are the rules around notification. Browsers are required to display an indicator that shows that a camera or microphone is in use, above and beyond any hardware indicator that may exist. They must also show an indicator that permission has been granted to use a device for input, even if the device is not actively recording at the moment.
For example in Firefox, the URL bar displays a pulsing red icon to indicate that recording is underway. The icon is gray if the permission is in place but recording is not currently underway. The device's physical light is used to indicate whether or not recording is currently active. If you've muted your camera (so-called "facemuting"), your camera's activity light goes out to indicate that the camera is not actively recording you, without discarding the permission to resume using the camera once muting is over.
There are a number of ways security management and controls in a user agent can cause getUserMedia() to return a security-related error.
Permissions Policy
The two Permissions Policy directives that apply to getUserMedia() are camera and microphone.
For example, this HTTP header will enable use of a camera by the document and any embedded <iframe> elements that are loaded from the same origin:
This will request access to the microphone for the current origin and the specific origin https://developer.mozilla.org:
If you're using getUserMedia() within an <iframe>, you can request permission just for that frame, which is clearly more secure than requesting a more general permission. Here, indicate we need the ability to use both camera and microphone:
<iframe src="https://mycode.example.net/etc" allow="camera; microphone">
</iframe>
Encryption based security
The getUserMedia() method is only available in secure contexts. A secure context is one the browser is reasonably confident contains a document which was loaded securely, using HTTPS/TLS, and has limited exposure to insecure contexts. If a document isn't loaded in a secure context, the navigator.mediaDevices property is undefined, making access to getUserMedia() impossible.
Attempting to access getUserMedia() in this situation will result in a TypeError.
Document source security
Because of the obvious security concern associated with getUserMedia() if used unexpectedly or without security being carefully managed, it can only be used in secure contexts. There are a number of insecure ways to load a document that might, in turn, attempt to call getUserMedia(). The following are examples of situations in which getUserMedia() is not permitted to be called:
- A document loaded into a sandboxed
<iframe> element cannot call getUserMedia() unless the <iframe> has its sandbox attribute set to allow-same-origin. - A document loaded using a
data:// or blob:// URL which has no origin (such as when one of these URLs is typed by the user into the address bar) cannot call getUserMedia(). These kinds of URLs loaded from JavaScript code inherit the script's permissions. - Any other situation in which there is no origin, such as when the
srcdoc attribute is used to specify the contents of a frame.
Generally, you will access the MediaDevices singleton object using navigator.mediaDevices, like this:
async function getMedia(constraints) {
let stream = null;
try {
stream = await navigator.mediaDevices.getUserMedia(constraints);
} catch (err) {
}
}
Similarly, using the raw promises directly, the code looks like this:
navigator.mediaDevices
.getUserMedia(constraints)
.then((stream) => {
})
.catch((err) => {
});
Note: If the current document isn't loaded securely, navigator.mediaDevices will be undefined, and you cannot use getUserMedia(). See Security for more information on this and other security issues related to using getUserMedia().
Below are some examples of the constraints parameter.
The following requests both audio and video without any specific requirements:
getUserMedia({
audio: true,
video: true,
});
While information about a user's cameras and microphones are inaccessible for privacy reasons, an application can request the camera and microphone capabilities it needs and wants, using additional constraints. The following expresses a preference for 1280x720 camera resolution:
getUserMedia({
audio: true,
video: { width: 1280, height: 720 },
});
The browser will try to honor this, but may return other resolutions if an exact match is not available, or the user overrides it.
To require a capability, use the keywords min, max, or exact (a.k.a. min === max). The following demands a minimum resolution of 1280x720:
getUserMedia({
audio: true,
video: {
width: { min: 1280 },
height: { min: 720 },
},
});
If no camera exists with this resolution or higher, then the returned promise will be rejected with OverconstrainedError, and the user will not be prompted.
The reason for the difference in behavior is that the keywords min, max, and exact are inherently mandatory — whereas plain values and a keyword called ideal are not. Here's a full example:
getUserMedia({
audio: true,
video: {
width: { min: 1024, ideal: 1280, max: 1920 },
height: { min: 576, ideal: 720, max: 1080 },
},
});
An ideal value, when used, has gravity — which means that the browser will try to find the setting (and camera, if you have more than one), with the smallest fitness distance from the ideal values given.
Plain values are inherently ideal, which means that the first of our resolution examples above could have been written like this:
getUserMedia({
audio: true,
video: {
width: { ideal: 1280 },
height: { ideal: 720 },
},
});
Not all constraints are numbers. For example, on mobile devices, the following will prefer the front camera (if one is available) over the rear one:
getUserMedia({
audio: true,
video: { facingMode: "user" },
});
To require the rear camera, use:
getUserMedia({
audio: true,
video: {
facingMode: { exact: "environment" },
},
});
Another non-number constraint is the deviceId constraint. If you have a deviceId from mediaDevices.enumerateDevices(), you can use it to request a specific device:
getUserMedia({
video: {
deviceId: myPreferredCameraDeviceId,
},
});
The above will return the camera you requested, or a different camera if that specific camera is no longer available. Again, to require the specific camera, you would use:
getUserMedia({
video: {
deviceId: {
exact: myExactCameraOrBustDeviceId,
},
},
});
This example gives a preference for camera resolution, and assigns the resulting MediaStream object to a video element.
const constraints = {
audio: true,
video: { width: 1280, height: 720 },
};
navigator.mediaDevices
.getUserMedia(constraints)
.then((mediaStream) => {
const video = document.querySelector("video");
video.srcObject = mediaStream;
video.onloadedmetadata = () => {
video.play();
};
})
.catch((err) => {
console.error(`${err.name}: ${err.message}`);
});
Lower frame-rates may be desirable in some cases, like WebRTC transmissions with bandwidth restrictions.
const constraints = {
video: { frameRate: { ideal: 10, max: 15 } },
};
On mobile phones.
let front = false;
document.getElementById("flip-button").onclick = () => {
front = !front;
};
const constraints = {
video: { facingMode: front ? "user" : "environment" },
};