Fully buffer video in Chrome

buffer, google-chrome, html, javascript, video

Solution

Because S3 supports partial downloads, Chrome initially buffers "only what's needed" in front of the playhead and then stops buffering. It will continue buffering "only what's needed" when the video starts playing. This means that depending on the user's download speed it will stop buffering now and then, just because it has enough buffer to play continuously.

But if you pause the video after having played some, Chrome will not stop buffering and go through all the way to the end.

This example exploits that technique to entirely buffer the video off screen before showing it on the page.

Tested on Chrome 32

// Create video in background and start playing
var video = document.createElement('video');
video.src = 'video.mp4';
video.controls = true;
video.muted = true; 
video.play();

// Pause immediately after it starts playing.
video.addEventListener("timeupdate", function() {
    if (this.currentTime > 0) {

        this.pause();
        video.muted = false;
        video.currentTime = 0
        this.removeEventListener("timeupdate", arguments.callee, false);

        // When whole video is buffered, show video on page
        video.addEventListener("progress", function() {
            if (Math.round(video.buffered.end(0)) / Math.round(video.seekable.end(0)) == 1) {
                document.body.appendChild(video);
                this.removeEventListener("progress", arguments.callee, false);
            }
        }, false);
    }
}, false);

Problem

Is it possible to fully buffer HTML5 video in Chrome (and Opera)? I host the movie in .mp4 and .webm on amazon S3. In HTML I use standard `<video>` tag. The server responds with status `206 Partial Content`. It is great, as it allows the browser to download any part of the video but I need to be able to seek instantly to any part of the file. I tried: - `.PROGRESS` event. When Chrome stops buffering the first part, the connection is killed. The next parts are downloaded in new connection, so JavaScript isn't able to continue checking total downloaded data. Even if it could, the old data isn't stored in cache. This method works in Firefox and Safari. Heck, even in IE10! - `XHRS`. I am able to fully download the movie, but when the video starts playing, Chrome tries to download the movie from the server again. I even tried to pass the movie into HTML in base64, but that's to much data - `FileAPI`. Chrome isn't able to create BLOB big enough and crashes. Any help is welcome! ps. I am aware of od and similar questions, but I hope something changed since they were asked.

Original source

Related problems