We just did the final touches of coding on a suite of desktop video workflow apps for a client – to automate the ingest of hundreds of Terabytes or Petabytes of media from professional video cameras, and import them quickly and reliably to a huge digital library.
Part of this suite is a scalable render farm using Adobe Media Encoder that we can put onto every machine in a facility, which can then activate overnight to churn through terabytes of processing while the humans sleep. Like the shoemaker’s elves.
As we’ve found with many MAM (media asset management) systems, the media import queue is too slow and crashy when importing lots of media, and requires pre-processing of files as they come out of professional cameras.
So the tools that we designed & built speed all this up by bypassing the MAM’s ingest systems – instead, we process everything on more powerful computers in parallel, then inject it all in the backdoor via an API.
Very geekily satisfying to conceive this kind of custom automated workflow and follow it all the way through to completion, knowing how much difference it will make.