We're in Bloomington, Indiana ready to record Derek Franklin, except for one small problem... The University won't allow us to shoot on campus. After wasting hours of calling around and trying to find the right people to talk to, we finally get into a room. Fantastic. I begin recording and about 20 minutes into the interview I notice a bit of text in the viewfinder "HDV-SD30". Uh, I know what HDV is, but what does the SD30 part mean? That typically means Standard Definition 30 frames per second, but that can't be right. As the interview is going on, I silently grab the camera manual and flip to the part about recording formats. Oh no. Turns out I AM right and HDV-SD30 means exactly what I thought it meant. We have to stop the recording (which was amazing thus far), and reset the camera to HDV-HD30 which is 720p 30 frames per second. Start the whole process again, about about 15 minutes into this take, the hair light blows out. This time I decide to keep recording and we finish the interview without any back lighting (which is why the image looks so flat). The last part of recording is to put in a second tape and record our host asking the questions again, allowing me to cut them in after the fact. Makes the interview flow a bit better when we can see who is asking the questions.

Get back to Minneapolis and begin capturing the footage. First tape captured fine, but the second tape had errors we were unable to get around. Such is the wonder of HDV, one small glitch and the whole shot is ruined (which is why professionals should not use HDV). Ok, no reverse interview shots. Finally finish editing and begin the h.264 conversion.

On my 1.8Ghz iMac G5 it took 6 days to compress all of the files from 480p to 1080p (although 1080p was 3 of the 6 days worth of encoding). This means that we have to be sure that the video is done right the first time. If I have to re-render it will take 6 more days to get it out the door. Fortunately we got the whole thing right the first time.

This is where the debate started, and where I would love your feedback: what is the best way to embed the video in the web site? Should we use the javascript popup like we are? Should we use a script that replaces the image with the video? Should we have the video play full-screen? Or should we just leave everything to RSS and have users syndicate everything? We started by posting the video as an embedded QuickTime file that would replace the image with the movie. If you have QuickTime 7 or better installed, you can see the example below:

The problem with the above example was that there was no way to allow the content to be seen in RSS syndication software, it just strips that part out. So you see the text of the story, but no video, the only way to get the video was to come to our site. While we love the visitors to the site, we want to make sure that everyone can get their daily dose of Technology Evangelist in their favorite program. So that basically eliminated the first option.

The second option is what you see now. Have the video pop in a javascript controlled window, as seen below. The disadvantage here is that it's not as sexy as the first option, and it can be blocked by pop-up blockers. You can see this link, and it will work in RSS readers such as Google Reader and Bloglines.

The third possibility is something we never posted, and that's called a .qtl file. This is basically an XML file that calls QuickTime player to open. Since this is linked to just like any other hyperlink, it will show up in bloglines and so forth. The other advantage is that I can control the player state by forcing the player full-screen or whatnot. I have yet to get this to work without crashing the QuickTime player, but you're free to try below. If we did this, I would then want to know if we should force the player full-screen, or should we just let the player open normally.

Standard Version -
Javascript Version -

This was the great debate, and to be honest we're still debating it. I think the best way to solve this is to have our own viewers tell us what does and does not work for them. Tomorrow I'll cover why we went with QuickTime (we actually didn't), and what equipment was used in the actual production.