Under Compression

  • Comments 0

Last week I was creating short introduction videos for our Architecture Guide project. You'd assume that this would be easy enough - write some slides and record the commentary, and then generate a WMV file from the recording. I used Camtasia, which integrates with PowerPoint and makes it really easy to create the recording and edit it. Only then, when I generated the WMV file, did I start to appreciate just how large these kinds of files can be.

You see, the problem is that we are limited to a maximum file size of 4MB and five or six minutes. Yet the first attempt using the default settings produced a 12MB file. So that's when I started digging around the settings, and reading up on video options and formats. You'd think that with the power and usability of modern software there would be some setting, like there is in Windows Movie Maker, saying "just make it fit into 4MB and be wonderful". No such luck.

Yes, I did try taking the WMV and exporting it from Movie Maker into 4MB, but the quality was so bad you couldn't even read the slide titles. So I tried recording in AVI format and converting that, with approximately equally awful results. In the end, I used the "Slides and Audio (Medium)" setting in Camtasia and edited the commentary (such as removing pauses and superfluous waffle) until it was just below 4MB. And then repeated the process for the other ten presentations on my "to do" list. The final quality is acceptable, though the audio compression makes me sound like I've developed a lithp.

Perhaps I just don't grasp the technicalities in enough depth. You can edit a whole range of audio and video parameters for the encoder, so in theory you can delicately tweak the settings to get an optimal output quality. But none of it seems to do quite what you'd expect. You reduce the frame rate from 10fps to 5fps and little dots start crawling all over the slides like a disorganized army of ants slowly eating the words. You change the keyframe interval from 1 second to 5 seconds, and you end up looking at the introduction slide for the first two minutes of the video. You reduce the audio bandwidth from 22KHz to 16KHz (mono) and it sounds like someone playing a kazoo.

It's rather like if your TV changed channel every time you adjusted the volume, or if the phone company rearranged the digits of your telephone number whenever someone tried to call you. I even worked diligently through a set of recommendations from some clever people who do this kind of thing all the time. It took ages to locate all the settings and make sense of them, and I wish I'd read the instructions from the end backwards rather than starting from step 1. When I finally got it all set up, I discovered the final paragraph of the document contained the rather less than useful comment "Using these settings, I managed to get the file size down to below 6MB per minute!" Wow, that's a lot of help when I need to get 5 or 6 minutes into 4MB...

Of course, after I finally did manage to squeeze my output into the required dimensions (if you'll pardon the expression) I thought it would be a good thing to document what I'd discovered. I mean, I work for the documentation team, so I probably ought to produce some dregs of documentation now and then - if only to justify my existence. Of course, in this brave new media-based world, it's actually "guidance" I create, not just boring old documentation, so I did think it would be neat to do a video on how to record videos. A sort of "guidance on guidance" thing. Only trouble is, I couldn't find a way to get Camtasia to allow me to record myself using Camtasia... There must be some way to do it, probably using a VM (or an ordinary video camera). Or by installing another video capture program to capture you using the original video capture program. I wonder if the guidance team at TechSmith (who make Camtasia) have to secretly smuggle somebody else's software onto their machines to create their guidance...?

Mind you, even better, next week I'm doing "train the trainer" videos to teach people how to teach people to use our gleaming new Architecture Guide (I bet you'd forgotten that this post was about the new guide). So if I create some documentation on how to do that, is it "guidance on guidance on guidance"? It's all starting to sound a bit like Chinese Whispers (or "Telephone" in the US). You know, the game they play at the kind of parties I'm never invited to where you have to pass on a whispered message and then see what it comes out as after ten or twenty people have communicated it.

I remember reading how, in the trenches during the First World War, supposedly they would pass commands back to the reserve lines in this way. Although it maybe won't make sense to any non-UK readers, the command "Send reinforcements, we're going to advance" is said to have been delivered to the reserve lines as "Send three and fourpence, we're going to a dance". At least it would probably be more entertaining than the videos I'm creating.

Leave a Comment
  • Please add 8 and 8 and type the answer here:
  • Post

Under Compression