What's important

Technically Speaking

There are a number of things which are really important to get right for your digital media.

While our aim is help you deliver your message without you having to worry about the technology behind it you may want to know how it all works or deliver some of it yourself.

Please look through the sections and see if they answer your questions. If not please contact us.






Normally the only post-processing that is required is

Sometimes the video is a little bit shaky at the start of the talk, for example if one speaker is quite small and the next quite tall. Image stabilisation tools only work to a certain extent - it handles the handheld shaky camera effect.
Stabalising a video like this requires lots of patience - you need to work on tiny segments of video and keep the speaker's head as steady as possible. Videoing the talk in widescreen and converting to 4:3 (on as blank as possible background) gives you most scope for this work.


You are starting at a high quality audio stream and encoding it to a low data rate (approximately 64Kbps) for the web. To ensure that you have the best quality we recommend a number of steps so that the compression starts from the important sound. Some of these need to be done a number of times - you may have to cycle through the steps


Encoding for the web is confusing. There is no standard format which can be played on all browsers. We therefore recommend encoding all of the following formats to ensure that your video can be played by everyone:

H.264 (.mp4), WebM (.webm)
.flv, Theora (.ogg) and .mp3


Our player initially tries to play the presentation in flash and if it can't play in flash (mobile phones, or flash not installed on a computer) it falls back to a HTML5 implementation. The reason for this is that our flash player uses streaming audio and video over the Akamai Network to ensure optimum delivery

The reason we initially try to stream using flash is that streaming lets you move to a portion of a talk which hasn't already downloaded. HTML5 can mimic it for video but for audio you need to download the talk before you can seek to a position. This is not a problem for a short talks but can be annoying for longer ones.

People watch 250% more content if they have an optimal video experience (fast startup, little to no buffering, and visual clarity). Read more.