Podcast Producer is an interesting technology that is only available with Mac OS X 10.5.x server software. One the surface, it is made to execute an unlimited amount of commands and technologies with the simple submission of your audio-only or video of your project. It can take care of placing titles and footers based off the title of your clip for you automatically, addressing multiple versions in differencing compression technologies, uploading to multiple destinations, creating and updating XML structures, e-mailing availability announcements, and more…with the right programming skills. That leads me to my next point…
As it is ridiculously easy to submit a project, it can be just potentially difficult to create these workflow if you don’t have the experience. A workflow is the set of commands you want executed on your media, in parallel. In a workflow, you can define the actions of multiple jobs to be done simultaneously before any compression take place. The compression jobs then can be executed, usually in parallel with one another. After which, the final pieces can be cataloged in XML and uploaded to any of the appropriate places.
Now Apple has made this relatively easy, considering you have to think in parallel for the entire job and you literally could have 10 things going on at once with in the workflow with another 15 waiting on the end condition of each of those 10. It still takes time to develop. Now, not as much as it would if you had to process 5 or 30 videos a week.
This is no small set up. In my experience, you need at least one Intel Mac with multiple cores running Mac OS X Server 10.5.x and at least one independent partition or drive dedicated to Podcast Producer. The optimal setup is multiple Intel Macs running Mac OS X or OS X Server 10.5.x. They need to use XGrid which is parallel processing software (built in to all versions of Mac OS X) and either shared space over Gigabit Ethernet or a XSan Fiber Channel System for shared storage. Podcast Producer only has to be on the controlling machine and it comes with OS X Server. If you want to do other formats than that of standard Quicktime, then the easiest is to spend $2,000 on a product from Telestream called Episode Podcast for Podcast Producer. As I understand it, you install this on the Podcast Producer Server and it will allow all of your machines in your XGrid to convert a myriad of formats, including MPEG-2, WMV, Flash.
My experience has been that this is a richly powerful system that was initially designed for the people who want to get a message out there but don’t have the skills to do 3 videos a week of prepping the title, closer, compression, xml documentation, and posting. This was designed for teachers and professors to get class recordings out for students in a timely fashion that can do a whole lot more.
For instance, I used this with a project I worked on in my days at the Cintas Center at Xavier University. The organization is called NDICE and you can see the 2008 talks here. We recorded roughly three 1 hour talks a day, in three days. I would crash digitize on a FCP system on my laptop. I would make sure to have a clean beginning and end. I would export without doing any titles or copyright info. I would then submit the file with the appropriate title, author, and timing information and let it rip. What I had Podcast Producer do is create an audio-only version and a video version with an opening title and closing copyright. For the audio-only version I wrote a little script to speak the title and place it at the head of the audio. For both clips I wrote another piece of code to do a custom insert into my XMLs for the website as well as place them at a location for me to move them to the web server. (I was using a different machine for my web server than my Podcast Producer Server and did not have the time to setup the transport script between the two). It worked well for our needs and when I had to change one title on a clip, all I had to do was resubmit of Podcast Producer and then after it was done do a little manual cleaning of the XML which wasn’t too bad. (The page uses the same XML as iTunes uses to generate its list) Just so you know, I never got the XGrid to run on more than one machine (I suspect my limited knowledge of Kerberos, and yes you have to use Kerberos and Open Directory only on Intel Macs) but it ran well on an 3.0 GHz 8-Core Mac Pro with 8 Gigs of RAM. Each Parallel process took up 3GHz, running 4 processes simultaneously. One more thing, in theory, you could write a piece of code to make sure that if the XML contained that clip, it would replace instead of appending.
Bottom line, it took me some time to come up with the code segments to do all of my customizations, but the infrastructure is sound to do it. And while you can’t submit a Motion Document or LiveType disguised as a Quicktime file (Believe me I tried since I STILL can not get Compressor to use a Cluster) you can take any other quicktime compatible file or anything else if you have the Episode Engine and use it to process to multiple formats. I have even come across people using lame and ffmpeg in XGrid to do all sorts of things. For large productions with multiple iterations requiring multiple outputs in different ways and locations, this is a dream…once it is configured correctly.
