Historically, writing scripts for the Web that work intimately with Web APIs has had a well-known challenge: often, your code needs to know whether or not an API exists and if so, what its limitations are on the 用户代理 it's running on. Figuring this out has often been difficult, and has usually involved looking at some combination of which 用户代理 (or browser) you're running on, which version it is, looking to see if certain objects exist, trying to see whether various things work or not and determining what errors occur, and so forth. The result has been a lot of very fragile code, or a reliance on libraries which figure this stuff out for you, then implement polyfills to patch the holes in the implementation on your behalf.
The twin concepts of constraints and capabilities let the browser and Web site or app exchange information about what constrainable properties the browser's implementation supports and what values it supports for each one. This article discusses capabilities and constraints, as well as media settings, and includes an example we call the Constraint Exerciser . The Constraint Exerciser lets you experiment with the results of different constraint sets being applied to the audio and video tracks coming from the computer's A/V input devices (such as its webcam and microphone).
The process works like this (using
MediaStreamTrack
as an example):
MediaDevices.getSupportedConstraints()
to get the list of
supported constraints
, which tells you what constrainable properties the browser knows about. This isn't always necessary, since any that aren't known will simply be ignored when you specify them—but if you have any that you can't get by without, you can start by checking to be sure they're on the list.
getCapabilities()
method; this object lists each supported constraint and the values or range of values which are supported.
applyConstraints()
method is called to configure the API as desired by specifying the values or ranges of values it wishes to use for any of the constrainable properties about which it has a preference.
getConstraints()
method returns the set of constraints passed into the most recent call to
applyConstraints()
. This may not represent the actual current state of the track, due to properties whose requested values had to be adjusted and because platform default values aren't represented. For a complete representation of the track's current configuration, use
getSettings()
.
In the Media Stream API, both
MediaStream
and
MediaStreamTrack
have constrainable properties.
If you need to know whether or not a given constriant is supported by the user agent, you can find out by calling
navigator.mediaDevices.getSupportedConstraints()
to get a list of the constrainable properties which the browser knows, like this:
let supported = navigator.mediaDevices.getSupportedConstraints();
document.getElementById("frameRateSlider").disabled = !supported["frameRate"];
In this example, the supported constraints are fetched, and a control that lets the user configure the frame rate is disabled if the
frameRate
constraint isn't supported.
A single constraint is an object whose name matches the constrainable property whose desired value or range of values is being specified. This object contains zero or more individual constraints, as well as an optional sub-object named
advanced
, which contains another set of zero or more constraints which the user agent must satisfy if at all possible. The user agent attempts to satisfy constraints in the order specified in the constraint set.
The most important thing to understand is that most constraints aren't requirements; instead, they're requests. There are exceptions, and we'll get to those shortly.
Most simply, each constraint may be a specific value indicating a desired value for the setting. For example:
let constraints = {
width: 1920,
height: 1080,
aspectRatio: 1.777777778
};
myTrack.applyConstraints(constraints);
In this case, the constraints indicate that any values are fine for nearly all properties, but that a standard high definition (HD) video size is desired, with the standard 16:9 aspect ratio. There's no guarantee that the resulting track will match any of these, but the user agent should do its best to match as many as possible.
The prioritization of the properties is simple: if two properties' requested values are mutually exclusive, then the one listed first in the constraint set will be used. As an example, if the browser running the code above couldn't provide a 1920x1080 track but could do 1920x900, then that's what would be provided.
Simple constraints like these, specifying a single value, are always treated as non-required. The user agent will try to provide what you request but will not guarantee that what you get will match. However, if you use simple values for properties when calling
MediaStreamTrack.applyConstraints()
, the request will always succeed, because these values will be considered a request, not a requirement.
Sometimes, any value within a range is acceptable for a property's value. You can specify ranges with either or both minimum and maximum values, and you can even specify an ideal value within the range, if you choose. If you provide an ideal value, the browser will try to get as close as possible to matching that value, given the other constraints specified.
let supports = navigator.mediaDevices.getSupportedConstraints();
if (!supports["width"] || !supports["height"] || !supports["frameRate"] || !supports["facingMode"]) {
// We're missing needed properties, so handle that error.
} else {
let constraints = {
width: { min: 640, ideal: 1920, max: 1920 },
height: { min: 400, ideal: 1080 },
aspectRatio: 1.777777778,
frameRate: { max: 30 },
facingMode: { exact: "user" }
};
myTrack.applyConstraints(constraints).then(function() => {
/* do stuff if constraints applied successfully */
}).catch(function(reason) {
/* failed to apply constraints; reason is why */
});
}
Here, after ensuring that the constrainable properties for which matches must be found are supported (
width
,
height
,
frameRate
，和
facingMode
), we set up constraints which request a width no smaller than 640 and no larger than 1920 (but preferably 1920), a height no smaller than 400 (but ideally 1080), an aspect ratio of 16:9 (1.777777778), and a frame rate no greater than 30 frames per second. In addition, the only acceptable input device is a camera facing the user (a "selfie cam"). If the
width
,
height
,
frameRate
，或
facingMode
constraints can't be met, the promise returned by
applyConstraints()
will be rejected.
Constraints which are specified using any or all of
max
,
min
，或
exact
are always treated as mandatory. If any constraint which uses one or more of those can't be met when calling
applyConstraints()
, the promise will be rejected.
So-called advanced constraints are created by adding an
advanced
property to the constraint set; this property's value is an array of additional constraint sets which are considered optional. There are few if any use cases for this feature, and there is some interest in removing it from the specification, so it will not be discussed here. If you wish to learn more, see
section 11 of the Media Capture and Streams specification
, past example 2.
可以调用
MediaStreamTrack.getCapabilities()
to get a list of all of the supported capabilities and the values or ranges of values which each one accepts on the current platform and user agent
.
This function returns a
MediaTrackCapabilities
object which lists each constrainable property supported by the browser and a value or range of values which are supported for each one of those properties.
注意：
getCapabilities()
hasn't been implemented yet by all major browsers. For the time being, you'll have to try to get what you need, and if you can't, decide what to do at that point. See Firefox
bug 1179084
，例如。
The first and most common way to use constraints is to specify them when you call
getUserMedia()
:
navigator.mediaDevices.getUserMedia({
video: {
width: { min: 640, ideal: 1920 },
height: { min: 400, ideal: 1080 },
aspectRatio: { ideal: 1.7777777778 }
},
audio: {
sampleSize: 16,
channelCount: 2
}
}).then(stream => {
videoElement.srcObject = stream;
}).catch(handleError);
In this example, constraints are applied at
getUserMedia()
time, asking for an ideal set of options with fallbacks for the video.
You can specify one or more media input device IDs to establish restrictions on which input sources are allowed. To collect a list of the available devices, you can call
navigator.mediaDevices.enumerateDevices()
, then for each device which meets the desired criteria, add its
deviceId
到
MediaConstraints
object that eventually gets passed into
getUserMedia()
.
You can also change the constraints of an existing
MediaStreamTrack
on the fly, by calling the track's
applyConstraints()
method, passing into it an object representing the constraints you wish to apply to the track:
videoTrack.applyConstraints({
width: 1920,
height: 1080
});
In this snippet, the video track referenced by
videoTrack
is updated so that its resolution as closely as possible matches 1920x1080 pixels (1080p high definition).
It's important to remember the difference between
constraints
and
settings
. Constraints are a way to specify what values you need, want, and are willing to accept for the various constrainable properties (as described in the documentation for
MediaTrackConstraints
), while settings are the actual values of each constrainable property at the current time.
If at any time you need to fetch the set of constraints that are currently applied to the media, you can get that information by calling
MediaStreamTrack.getConstraints()
, as shown in the example below.
function switchCameras(track, camera) {
let constraints = track.getConstraints();
constraints.facingMode = camera;
track.applyConstraints(constraints);
}
This function accepts a
MediaStreamTrack
and a string indicating the camera facing mode to use, fetches the current constraints, sets the value of the
MediaTrackConstraints.facingMode
to the specified value, then applies the updated constraint set.
Unless you only use exact constraints (which is pretty restrictive, so be sure you mean it!), there's no guarantee exactly what you're going to actually get after the constraints are applied. The values of the constrainable properties as they actually are in the resulting media are referred to as the settings. If you need to know the true format and other properties of the media, you can obtain those settings by calling
MediaStreamTrack.getSettings()
. This returns an object based on the dictionary
MediaTrackSettings
。例如：
function whichCamera(track) {
return track.getSettings().facingMode;
}
此函数使用
getSettings()
to obtain the track's currently in-use values for the constrainable properties and returns the value of
facingMode
.
In this example, we create an exerciser which lets you experiment with media constraints by editing the source code describing the constraint sets for audio and video tracks. You can then apply those changes and see the result, including both what the stream looks like and what the actual media settings are set to after applying the new constraints.
The HTML and CSS for this example are pretty simple, and aren't shown here. You can look at the complete example by clicking here .
<p>Experiment with media constraints! Edit the constraint sets for the
video and audio tracks in the edit boxes on the left, then click the
"Apply Constraints" button to try them out. The actual settings the
browser selected and is using are shown in the boxes on the right.
Below all of that, you'll see the video itself.</p>
<p>Click the "Start" button to begin.</p>
<h3>Constrainable properties available:</h3>
<ul id="supportedConstraints">
</ul>
<div id="startButton" class="button">
开始
</div>
<div class="wrapper">
<div class="trackrow">
<div class="leftside">
<h3>Requested video constraints:</h3>
<textarea id="videoConstraintEditor" cols=32 rows=8></textarea>
</div>
<div class="rightside">
<h3>Actual video settings:</h3>
<textarea id="videoSettingsText" cols=32 rows=8 disabled></textarea>
</div>
</div>
<div class="trackrow">
<div class="leftside">
<h3>Requested audio constraints:</h3>
<textarea id="audioConstraintEditor" cols=32 rows=8></textarea>
</div>
<div class="rightside">
<h3>Actual audio settings:</h3>
<textarea id="audioSettingsText" cols=32 rows=8 disabled></textarea>
</div>
</div>
<div class="button" id="applyButton">
Apply Constraints
</div>
</div>
<video id="video" autoplay></video>
<div class="button" id="stopButton">
Stop Video
</div>
<div id="log">
</div>
body {
font: 14px "Open Sans", "Arial", sans-serif;
}
video {
margin-top: 20px;
border: 1px solid black;
}
.button {
cursor: pointer;
width: 150px;
border: 1px solid black;
font-size: 16px;
text-align: center;
padding-top: 2px;
padding-bottom: 4px;
color: white;
background-color: darkgreen;
}
.wrapper {
margin-bottom: 10px;
width: 600px;
}
.trackrow {
height: 200px;
}
.leftside {
float: left;
width: calc(calc(100%/2) - 10px);
}
.rightside {
float: right;
width: calc(calc(100%/2) - 10px);
}
textarea {
padding: 8px;
}
h3 {
margin-bottom: 3px;
}
#supportedConstraints {
column-count: 2;
-moz-column-count: 2;
}
#log {
padding-top: 10px;
}
Now let's take a look at the JavaScript code that makes everything work.
First we have the default constraint sets, as strings. These strings are presented in editable
<textarea>
s, but this is the initial configuration of the stream.
let videoDefaultConstraintString = '{\n "width": 320,\n "height": 240,\n "frameRate": 30\n}';
let audioDefaultConstraintString = '{\n "sampleSize": 16,\n "channelCount": 2,\n "echoCancellation": false\n}';
These defaults ask for a pretty common camera configuration, but don't insist on any property being of special importance. The browser should do its best to match these settings but will settle for anything it considers a close match.
Then we initialize the variables which will hold the
MediaTrackConstraints
objects for the video and audio tracks, as well as the variables which will hold references to the video and audio tracks themselves, to
null
.
let videoConstraints = null; let audioConstraints = null; let audioTrack = null; let videoTrack = null;
And we get references to all of the elements we'll need to access.
let videoElement = document.getElementById("video");
let logElement = document.getElementById("log");
let supportedConstraintList = document.getElementById("supportedConstraints");
let videoConstraintEditor = document.getElementById("videoConstraintEditor");
let audioConstraintEditor = document.getElementById("audioConstraintEditor");
let videoSettingsText = document.getElementById("videoSettingsText");
let audioSettingsText = document.getElementById("audioSettingsText");
These elements are:
videoElement
<video>
element that will show the stream.
logElement
<div>
into which any error messages or other log-type output will be written.
supportedConstraintList
<ul>
(unordered list) into which we programatically add the names of each of the constrainable properties supported by the user's browser.
videoConstraintEditor
<textarea>
element that lets the user edit the code for the video track's constraint set.
audioConstraintEditor
<textarea>
element that lets the user edit the code for the audio track's constraint set.
videoSettingsText
<textarea>
(which is always disabled) that displays the current settings for the video track's constrainable properties.
audioSettingsText
<textarea>
(which is always disabled) that displays the current settings for the audio track's constrainable properties.
Finally, we set the current contents of the two constraint set editor elements to the defaults.
videoConstraintEditor.value = videoDefaultConstraintString; audioConstraintEditor.value = audioDefaultConstraintString;
To the right of each of the constraint set editors is a second text box which we use to display the current configuration of the track's configurable properties. This display is updated by the function
getCurrentSettings()
, which gets the current settings for the audio and video tracks and inserts the corresponding code into the tracks' settings display boxes by setting their
值
.
function getCurrentSettings() {
if (videoTrack) {
videoSettingsText.value = JSON.stringify(videoTrack.getSettings(), null, 2);
}
if (audioTrack) {
audioSettingsText.value = JSON.stringify(audioTrack.getSettings(), null, 2);
}
}
This gets called after the stream first starts up, as well as any time we've applied updated constraints, as you'll see below.
buildConstraints()
function builds the
MediaTrackConstraints
objects for the audio and video tracks using the code in the two tracks' constraint set edit boxes.
function buildConstraints() {
try {
videoConstraints = JSON.parse(videoConstraintEditor.value);
audioConstraints = JSON.parse(audioConstraintEditor.value);
} catch(error) {
handleError(error);
}
}
This uses
JSON.parse()
to parse the code in each editor into an object. If either call to JSON.parse() throws an exception,
handleError()
is called to output the error message to the log.
startVideo()
method handles setting up and starting the video stream.
function startVideo() {
buildConstraints();
navigator.mediaDevices.getUserMedia({
video: videoConstraints,
audio: audioConstraints
}).then(function(stream) {
let audioTracks = stream.getAudioTracks();
let videoTracks = stream.getVideoTracks();
videoElement.srcObject = stream;
if (audioTracks.length) {
audioTrack = audioTracks[0];
}
if (videoTracks.length) {
videoTrack = videoTracks[0];
}
}).then(function() {
return new Promise(function(resolve) {
videoElement.onloadedmetadata = resolve;
});
}).then(function() {
getCurrentSettings();
}).catch(handleError);
}
There are several steps here:
buildConstraints()
to create the
MediaTrackConstraints
objects for the two tracks from the code in the edit boxes.
navigator.mediaDevices.getUserMedia()
, passing in the constraints objects for the video and audio tracks. This returns a
MediaStream
with the audio and video from a source matching the inputs (typically a webcam, although if you provide the right constraints you can get media from other sources).
<video>
element so that it's visible on screen, and we grab the audio track and video track into the variables
audioTrack
and
videoTrack
.
onloadedmetadata
event occurs on the video element.
getCurrentSettings()
function (described above) to display the actual settings that the browser decided upon after considering our constraints and the capabilities of the hardware.
handleError()
method that we'll look at farther down in this article.
We also need to set up an event listener to watch for the "Start Video" button to be clicked:
document.getElementById("startButton").addEventListener("click", function() {
startVideo();
}, false);
Next, we set up an event listener for the "Apply Constraints" button. If it's clicked and there's not already media in use, we call
startVideo()
, and let that function handle starting the stream with the specified settings in place. Otherwise, we follow these steps to apply the updated constraints to the already-active stream:
buildConstraints()
is called to construct updated
MediaTrackConstraints
objects for the audio track (
audioConstraints
) and the video track (
videoConstraints
).
MediaStreamTrack.applyConstraints()
is called on the video track (if there is one) to apply the new
videoConstraints
. If this succeeds, the contents of the video track's current settings box are updated based on the result of calling its
getSettings()
方法。
applyConstraints()
is called on the audio track (if there is one) to apply the new audio constraints. If this succeeds, the contents of the audio track's current settings box are updated based on the result of calling its
getSettings()
方法。
handleError()
is used to output a message into the log.
document.getElementById("applyButton").addEventListener("click", function() {
if (!videoTrack && !audioTrack) {
startVideo();
} else {
buildConstraints();
if (videoTrack) {
videoTrack.applyConstraints(videoConstraints).then(function() {
videoSettingsText.value = JSON.stringify(videoTrack.getSettings(), null, 2);
}).catch(handleError);
}
if (audioTrack) {
audioTrack.applyConstraints(audioConstraints).then(function() {
audioSettingsText.value = JSON.stringify(audioTrack.getSettings(), null, 2);
}).catch(handleError);
}
}
}, false);
Then we set up the handler for the stop button.
document.getElementById("stopButton").addEventListener("click", function() {
if (videoTrack) {
videoTrack.stop();
}
if (audioTrack) {
audioTrack.stop();
}
videoTrack = audioTrack = null;
videoElement.srcObject = null;
});
This simply stops the active tracks, sets the
videoTrack
and
audioTrack
变量到
null
so we know they're gone, and removes the stream from the
<video>
element by setting
HTMLMediaElement.srcObject
to
null
.
This code adds simple support for tabs to the
<textarea>
elements by making the tab key insert two space characters when either constraint edit box is focused.
function keyDownHandler(event) {
if (event.key == "Tab") {
let elem = event.target;
let str = elem.value;
let position = elem.selectionStart;
let newStr = str.substring(0, position) + " " +
str.substring(position, str.length);
elem.value = newStr;
elem.selectionStart = elem.selectionEnd = position + 2;
event.preventDefault();
}
}
videoConstraintEditor.addEventListener("keydown", keyDownHandler, false);
audioConstraintEditor.addEventListener("keydown", keyDownHandler, false);
The last significant piece of the puzzle: code that displays, for the user's reference, a list of the constrainable properties which their browser supports. Each property is a link to its documentation on MDN for the user's convenience. See the 范例 in MediaDevices.getSupportedConstraints() for details on how this code works.
Of course, there may be non-standard properties in this list, in which case you probably will find that the documentation link doesn't help much.
let supportedConstraints = navigator.mediaDevices.getSupportedConstraints();
for (let constraint in supportedConstraints) {
if (supportedConstraints.hasOwnProperty(constraint)) {
let elem = document.createElement("li");
elem.innerHTML = "<code><a href='https://developer.mozilla.org/docs/Web/API/MediaTrackSupportedConstraints/"
.concat(constraint) + "' target='_blank'>" + constraint + "</a></code>";
supportedConstraintList.appendChild(elem);
}
}
We also have some simple error handling code;
handleError()
is called to handle promises which fail, and the
log()
function appends the error message to a special logging
<div>
box under the video.
function log(msg) {
logElement.innerHTML += (msg + "<br>");
}
function handleError(reason) {
log("Error <code>" + reason.name +
"</code> in constraint <code>" + reason.constraint +
"</code>: " + reason.message);
}
Here you can see the complete example in action.
| 规范 | 状态 | 注释 |
|---|---|---|
|
媒体捕获和流
The definition of 'Constrainable pattern' in that specification. |
候选推荐 | 初始定义。 |
MediaDevices.getSupportedConstraints
| 桌面 | 移动 | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
getSupportedConstraints
|
Chrome 53 | Edge 12 | Firefox 44 | IE 不支持 No | Opera 40 | Safari 11 | WebView Android 53 | Chrome Android 52 | Firefox Android 50 | Opera Android 41 | Safari iOS 11 | Samsung Internet Android 6.0 |
完整支持
不支持
MediaTrackCapabilities
MediaTrackConstraints
MediaTrackSettings
MediaDevices.getSupportedConstraints()
MediaStreamTrack.applyConstraints()
MediaStreamTrack.getSettings()
AudioStreamTrack
BlobEvent
CanvasCaptureMediaStream
MediaDevices
MediaStream
MediaStreamTrack
MediaStreamTrackEvent
MediaTrackCapabilities
MediaTrackConstraints
MediaTrackSettings
MediaTrackSupportedConstraints
NavigatorUserMedia
NavigatorUserMediaError
VideoStreamTrack
DoubleRange
ConstrainDouble
LongRange
ConstrainLong
ConstrainBoolean
ConstrainDOMString