After the heinous webinar I did via LiveMeeting a few weeks ago, I've finally managed to successfully record the webinar so that people can actually see and hear the demos. It was a surprisingly large PITA for something that I would think "everyone" does. Maybe I just don't hang with the right crowd.
I first tried a product called Captivate to record the PowerPoint and demos. Some of our marketing guys had used it before with success. It generally worked ok, until it came time to demonstrate some commandline tools via RDP. Captivate would show the commandline window, and play little clickety-click noises when I typed, but it would NOT display anything that I typed until long after the fact. It made it really hard to understand what I was doing. You'd see the commandline prompt, here some clickety-click key-noises (please, is that REALLY necessary?), then a whole bunch of text would scroll by, with no clue as to what I had typed.
I downloaded and tried two or three other products that purported to do full screen capture videos, but the image quality was unbelievably poor. I mean REALLY poor. You couldn't decipher the text in the commandline window at 1024x768.
I finally tried Microsoft's Media Encoder, which is a free download. After extended fiddling with various settings (look, I know squat about MGEG-4 and AAC and stuff like that), I finally managed to get a screen capture that looked good, sounded good, and I could include in the PowerPoint presentation. The only problem? The resulting files were HUGE, like hundreds of MBs for a 10 minute demo. And there was a problem with the sound after all... when I "previewed" the recording, it sounded great. When I saved the recording and played it back later, the sound was great for the first 30 seconds or so, and then the pitch of my voice started to change up and down. Do you remember when you were a kid and you played vinyl LPs on a record plater, and spun the platter around by hand to make it speed up and slow down? That's what it sounded like... it was my voice, only with a whammy-bar.
I spent the next few days trying to sort out what the problem with the audio was. Sometimes it would get better, sometimes worse. After reading through the docs for WME again, I discovered that the recommended hardware for WME included a dual proc CPU. And I was recording on my HP NC6000 with a 2GHz P4M. So I took one of my 3.2Ghz test servers out of production and tried recording on that, and voila!, it worked! Note to self: Requisition a new dual-core laptop, like the Dell M1710. Is that a result of the AlienWare acquisition? Sure looks like it.
But I still had the problem with the large files. As it turns out, WME can compress media files to varying levels. I tried a couple of different compression settings, and got something that looked and sounded it pretty good. So now our web guy Henry is busy encoding the files and will combine them into some sort of downloadable package, maybe in Flash or some such. Look for an announcement as soon as he gets it done and posts the files.