Video Encoding for the Web
The internet is a very cost-effective method of distribution for your final project, whether it had been intended as a Web Video Production or not. But the video must first be optimized for online delivery to ensure a smooth playback, and to minimize your website's bandwidth costs. This is accomplished through a process known as Encoding.
There are two elements that need to be considered: The required File Format(s), and the Target File Size or Bitrate.
File Format
There are three main video formats currently considered to be ‘standard’ for web delivery: Flash, Quicktime and Windows Media. Each have strengths and weaknesses, but it really comes down to what your target audience is using.
This chart shows which of these formats come pre-installed on Mac and Windows machines, and the overall market penetration of the 3 formats.
Flash has become the format of choice for websites such as YouTube. Although Quicktime offers the highest quality in the smallest file size, it only comes pre-installed on Macs. Windows Media offers a large pre-installed user base, however this comes at the expense of picture and sound quality.
Often a quick conversation with your web designer can greatly aid you in making this decision.
Target File Size or Bitrate
Let’s look at a typical scenario: Your finished video is 5 minutes long. At full resolution (TV quality) your video will be about 900MB – far too large to put online! Ideally, we'd like to get your video closer to 9MB while still maintaining an acceptable level of quality in picture and sound.
The process begins by determining the Target File Size, conforming all other variables to suit. Inversely, a Target Bitrate may be established which is a factor of filesize per second of video. Multiplying this Bitrate by the number of seconds in the video will determine the File Size.
This is where the real work begins, as the Frame Size is first reduced, and the video is compressed.
Frame Size
The first step is to reduce the Frame Size. Full-quality Video is 640×480 pixels. In reducing this to between 320×240 to 480×360 pixels, we can reduce the File Size significantly. In some cases, the required Frame Size will be determined by the website in which the video will be played.
Compression
Powerful Hardware and Software are used to examine each frame of the video, essentially looking for patterns between pixels both within a single frame and between adjacent frames of video. This is accomplished by using a mathematical algorithm known as a CODEC (short for compression/decompression), and through two methods of compression known as Temporal and Spatial. The video is then decompressed by the viewer's computer after it has been downloaded in its compressed form.
In the Temporal method of compression, the computer is looking for instances where information remains relatively the same over time. For instance, if your video consists of a person standing in front of a simple white wall, the software is able to take a snapshot of the background area and repeat it in subsequent frames (until the background changes significantly).
Spatial Compression seeks patterns within a frame. Again with our white wall as an example, instead of each pixel of white having to be 'described' individually, the computer can group these pixels together and call them all 'white'.
In Summary
This may still all sound a little confusing to you, and that's okay. We'll understand if you just want your video to look 'good'. Contact us and we'll be more than happy to talk you through it.