In my professional life, I’m often tasked with making a help system of somekind to help users understand the application I’m building. This is notiously difficult to do. However, certain technologies have come along to make showing users what to do possible, even over dialup connections. One of those technologies is Flash, which combinds lightweight vector graphics with mp3 support.

Unfortunately, the Flash studio app costs mucho dinero and I wouldn’t use the tool enough to get the full value of my investment back. Also, what I want to do is capture stuff I’m doing on my desktop along with a voiceover of what’s going on. This, to my knowledge, would require an additional package.

Fortunately, there have been a few open source projects that can manipulate the Flash file format to some degree. But what I need is to capture a Windows desktop session.

The solution comes in the python version of vnc2swf. This appears to be a platform-neutral solution, which is remarkable.

Very. Good. Hack.

I wish perl had done this first, but hats off to the python hackers that made this happen. I applaud the use of a scripting language I can hack. I particularly like the edit.py feature, which allows me to add an mp3 soundtrack. I have a lot of equipment and software to manipulate sound, including Sonar for the base recording and WAV editting and RazorLame for the mp3 encoding (which I did at 32kb sample rate monophonically). I did have to time shift the original VO track to 85% of the original. That’s why the demo sounds like I just snorted crystal meth. I’m not sure if there’s a better trick for syncing the video and audio.

Click on the picture for the full demo of my wonderful site.