Video Compression Explained, Or How To Squash An Elephant

I wanted to send an elephant to a friend.  The postage was pretty severe, so I got a life size picture of one instead.  This was almost as good as the real thing, but it wouldn’t fit through her letterbox.  So I got a postcard made up.

This is about the best metaphor I can come up with for the video workflow in terms of data management.  We film in as high an image quality as we can afford, which creates footage that is malleable with little degradation.  But high quality also means high data rates, (the flow of data through your computer as the video plays).  If you’re filming for non-broadcast use a typical data rate would be around 28Mb per second.  Thus video editing suites need to be very high powered computers which are able to manipulate this material without crashing.

But if we delivered the final product to you in the same format we filmed it in it wouldn’t be fit for purpose.  Most office pcs would struggle if you played back original footage, and if you tried to stream it – watch it through the web or intranet – no hope.  That’s why we compress video before delivery.

Compression is when we make our elephant into a postcard.

The function of video compression is to make the file size and data rate suitable for the viewing platform.  There are two important factors here:

  • image size
  • codec

When you deliver video for e-learning it usually doesn’t play full-screen, so this reduction of the image size naturally presents an opportunity to reduce the file size and data rate of a clip.

A codec is software with which the video is re-encoded.  The software uses clever maths to take away as much of information it can and still make the image look acceptable, thus reducing file size and the data rate of playback.

Having reduced the image size and re-encoded the footage instead of needing to deliver 25 life size images a second we now have 25 postcards.

I feel obliged to add that there are also container file formats such as Flash Video and mpeg4 – which we can think of as the suffix or the file type.  These are effectively wrappers for the video file which can be encoded in one of several codecs.  And there are even file formats that are wrapper and codec in one.  I suggest you focus on the basic idea of video compression and glide over this detail – remembering that there’s some fiddly stuff that those tech guys need to deal with.

Different codecs and file formats use different means of compression and the resulting files have different qualities making them suitable for different uses.

To give one example – the flv (Flash Video) format from Adobe offers excellent options for high quality with low data rate and file size, but Apple does’t like Adobe products.  So you wouldn’t use it to encode video for tablets as the iPad dominates the market.

Everyone involved in the production of the video content should have an idea of the final video image size and quality, as well as knowledge of any interface in which the clips will appear.  The video content should be honed to these ends.  So, if you are going to play back the video as a very small image size of low quality you’d be wise to film bold compositions without too much detail.

A final word about HTML5:  To embed video into platforms created in HTML5 you need to encode the video in a variety of file types, each using a different codec and you must host all the versions.  The correct file type will be chosen for playback depending on the device and browser it’s being viewed on.

And that’s how to squash an elephant.

Share on linkedin
Share on twitter
Share on facebook
Share on skype
Share on email

Want to make change happen together?​