Podcast Producer Experience

This is all sort of unrelated to most of my other posts but I tried my first semi-official video podcast recording today. This is part of setting up infrastructure in the BARC (Barrow Arctic Research Center) for researchers to use while in the field. I had set up Podcast Producer 2 that is part of Apple OS X server and was having a lot of trouble with authentication. I shut down the Open Directory and the Podcast Producer services, deleted the databases associated with both services, and started over. After that, it worked flawlessly. I think it got messed up since I first installed Leopard Server then upgraded it to Snow Leopard. It didn't transition well. I also recorded the presentation in case I wanted to play with it later and this is the setup I used. Podcast Capture used the external iSight camera mounted to the DV camera and I took a feed from the audio mixer as my audio input. To do this sort of thing in "production", I see a need for at least 3 video feeds. One of the speaker, one from whatever is being sent to the projector and one of the audience. You can't hear or see the questions. Unfortunately, this then becomes a much bigger production. I'll need to think about it. The speaker didn't want me to publicize this video until he gets permission from his company but it was a fascinating talk on off-shore drilling platforms. The talk ended up being 1½ hours long and should be edited but I was trying to completely automate the post production process. The Xgrid is still chunking away encoding the video taking much longer than real time. I guess I need a bigger Xgrid (there is only one single processor machine in the Xgrid since this is an experiment) but it is only running at 20% of capacity. Maybe there are some controls to limit load. This is all new to me.

It is really windy today something like 26 knots. Blowing snow, lots of drifts and visibility is way down. It was a challenge just driving the truck from the BARC to NARL building 360 due to the low visibility. I suspect that my flight out tonight will get canceled.

Sunday evening - By the time I got back home, the encoding was done. As you can see from this graph of server utilization, it looks like it took almost 8 hours to encode the video. You can also see that the encoding was only being done on a single core of the server. I guess this is by design so that multiple jobs could be done simultaneously. But I have to admit that I wish it would've parallelized (is that a word?) the task. The 1½ hour recording ended up being almost 1 GB in size for the video and 100 mB for the audio. Wrong encoder choice I guess. I'll look at my selections tomorrow. Don't need stereo, 15 frames per second of less for a talking head, maybe a little editing of the file....