I want to display captions using the HTML5 Text Tracks API (WebVTT) using the native browser implementations -- without loading audio/video.
How can that be achieved ?
There is no way using the native implementation.
However, the Firefox implementation of WebVTT, vtt.js, is publicly available.
All you need to do is attach a timer while a video element is playing (or just to any custom event without a video) and apply vtt.js functions to the video / captions container.
Since the Firefox VTT implementation is derived from it, this is the closest you can come to native without using the default way of attaching to a video.
Related
I assume that this is possible, here is question, also on stackoverflow, that confirms it.
However I have failed while trying to make it work, so I'm asking for your help.
How can I do this?
WebM uses a variant of the Matroska container so it's not too surprising that some browsers are able to play some .mkv files. It's likely that you'd need to set the content type header to something that the browser recognises (eg. video/webm) if it was going to work at all.
However, that's only part of the story. The Matroska container is just that, a container. Inside the container are video and audio objects which have been encoded with particular codecs, if the browser you're attempting to play the file with doesn't support those codecs then they won't be able to play the file even if they can parse the container format. If you want wide compatibility then your best option is to transcode to one of the more common container/codec combinations such as MP4/h264 or WebM/VP8.
You wanna take a look at DivX player
Cut & paste
As of version 2.1, the DivX Plus Web Player supports the HTML5 API, so
it can be embedded and controlled using the tag and open web
standards such as JavaScript. It takes full advantage of the
tag’s attributes by allowing publishers to create seamless playback
experiences and custom controls for DivX, AVI, MKV, MP4 and MOV videos
across multiple browsers: Firefox and Chrome (Internet Explorer &
Safari Coming in full release version).
I'm working with a DirectShow application that exports as a Firefox plugin for video streaming.
To the best of my knowledge, it seems as though the plugin is being rendered on a separate window from the web page. I'd like to edit the application to allow html elements over the plugin much like flash allows with wmode='opaque' or 'transparent'. Another option that would work is to pass mouse events through the application (such as mouse move and mouse down).
I've been reading up on put_MessageDrain and put_Owner, but I'm not sure how to apply these to the application or if they will work.
I'm also aware that iframe shimming may work, but since I have access to the source code of the application I'd like to add the functionality directly rather than use a work around. Is iframe shimming my best option? It seems like the easiest option.
The application currently uses Windowless mode and is always on top of everything else on the web page.
The way video is displayed with DirectShow video renderers, you cannot make it transparent without providing your own video renderer filter, or custom allocator/presenter object for one of the standard video renderers. DirectShow renderers display video with DirectDraw or Direct3D surface which stands alone from hosting window and is not interacting with host in a way supposed to get you combined picture with transparent video and something else as well.
I've been reading up on put_MessageDrain and put_Owner
These are for Windowed mode, you seem to use Windowless which assumes that you are providing a hosting window and rectangle to cover opaquely. Renderless, the third mode VMR-7/9 and EVR offer you is the one where you provide allocator/presenter and override the presentation.
Windows SDK (for some you might need to look for an older version, not the latest) provides with samples: Renderless, VMR9Allocator, EVRPresenter which demo the renderless mode.
How does one go about streaming video in HTML5 ? I can go with using a single browser of the latest version if I have to.I need to be able to start playing from any location of the movie even if the entire video has not been loaded by the browser.
WebRTC ?
I've already seen this question and no one has answered.
Does not allow the viewer to skip to the middle of a video in any
browser. They must watch the video straight through start to finish,
which is not ideal.
This is the main point for streaming.
Currently if you want to use pure HTML5 and work cross-browser you are limited to progressive streaming with the <video> element.
While that still does allow the user to skip ahead via the scrubber or programatically by setting the .currentTime there will still be some buffering while the browser re-loads enough content to be comfortable playing smoothly.
Solutions like Smooth Streaming, HLS do not work across browsers today so you would require a Flash or Silverlight plugin, though with MPEG-DASH being recognized by the W3C there is some hope for the future as samples like this demonstrate http://dash-mse-test.appspot.com/release-notes.html
For today however if you want to stick with an HTML5 solution and you have source in a format the works with the browser then you should be fine
Chrome has implemented the Media Source API in the mean time. Hoping the rest will follow.
http://www.w3.org/TR/media-source/
Abstract
This specification extends HTMLMediaElement to allow JavaScript to generate media streams for playback. Allowing JavaScript to generate streams facilitates a variety of use cases like adaptive streaming and time shifting live streams.
I want to create a html5 page with video and an image overlay - meaning some image that is showing over the video. This overlay will in time also be text in some cases. Is there any good way to achieve this?
What I've been trying this far is to use a <video> tag to hold the video, and draw the image into a canvas, which I place on top of the video. To show it I need to move the video back setting z-index to -1, but then the video controls won't work. Maybe there's a solution to make the controls work again, but I'm not sure if I'm on the right path here.. I am assuming there is a recommended solution to this. Maybe using a canvas which I fill both video and overlay into. Or something completely different?
Note: I edited the question as it originally pointed in the wrong direction regarding what was important here. I'd love to have a solution which makes this work seamlessly in fullscreen and everything, but the focus is: What is the appropriate way to place items on top of video - in html5?
Achieving what you want and have it supported in out-of-the-box fullscreen is problematic. Fullscreen support in html5 video is only optional and in any way not accesible thorugh the API (See discussion here).
Even if you used the built in fullscreen there is no way you could inject content above it unless you are willing to change the video file itself on the server in runtime.
what you can do however (And what I did in a similar case) is to implement your own video controls, run the video tag without the built in controls, and have fun with overlaying as many layers as you want on top of your now out of focus video.
As for fullscreen, you can implement some sort of custom background fullscreen similar to what's been done here
edit: The problem you're having by placing a canvas over the video is blocking the built in html video controls. My suggestion is to implement your own video controls (play, pause, volume, seeker, etc.) using html and javascript calling the video API. You can probably even make it prettier then the ugly built in controls.
Your controls can be contained in a layer above the overlaid canvas, and thus the video will be shown, above it the overlay and above it your control set.
You can read a little about implementing your own controls here or here
And anyway this can easily be much better than this.
I'm trying to write a simple audio player for a website, and am using the EMBED... tag to embed the audio and setting HIDDEN="true" and using various javascript commands to control the audio playback. It works fine for realplayer and mplayer but the quicktime plugin doesn't respond to javascript if the hidden bit is set - is there any workaround for this?
First, i suggest you to use the object html tag which is standardized, embed is not.
Then you could simply hide your embeded audio using CSS instead of this hidden param.
Even better, you should hide it using CSS through JavaScript, because if you do it this way, people who don't have JavaScript enabled but support CSS are able use the plugin player directly.
I found that setting height=0 width=0 worked the same as hidden=true and solved the problem
Actually, the Quicktime plugin requires that at least a 12px squared area be shown, otherwise it will not load, and the JavaScript API will not be exposed. The area detection is done at short intervals, and if the visible area is ever reduced, the playback is stopped immediately. The plugin is extremely good at determining if it can actually be seen, you can't event cover it with other elements.