Putting aside the mangled Shakespeare reference, this is one of the questions where the rubber meets the road in the design efforts for a DAM between a good engineering solution and a good business solution.
Here's the rub. A good engineering solution would be to create the highest quality master files possible of the physical assets and then create; via transcode; the files needed for further processing or delivery purposes on-demand. This is an efficient way to manage storage and transcode resources while slowly building a library of sub-files for assets. This makes sense but requires a robust transcode and storage management solution to handle the load on-demand within the service level agreement (SLA) required by the business needs and marketplace. For example, it just may not be acceptable to have to wait (even a few minutes) for a file creation if your business needs dictate that the files are available immediately for download. This is especially critical with large files in a media or video environment and in situations where lots of different departments are tapping into the same transcode resources for this work.
A good business solution is to define the file types most commonly used for the business and transcode in advance as part of the ingest process so that upfront all the materials needed are readily available. Basically, pre-populate so the items are there when needed. I like to call these House Flavors and if the planning is done right, you should be able to build a collection of assets that all have the same flavors of files at the ready. With this type of solution you still need to have the ability to make the custom file types as needed on-demand but these become the exceptions and you can tune your infrastructure to the predictable loads that come from a pre-determined house flavor file set and the production process to create them. You also need to plan for the additional storage that this would create the need for and the storage management for managing your storage efficiently. In this model, the focus shifts to outputting large amounts of content quickly not creating them first.
Neither solution is exclusively right or wrong, they just require different types of trade-offs and support. You have to change the conditions and the infrastructure to support the path chosen. The bottom-line is that ideally technologies should remain invisible to the business users. It shouldn't be an obstacle to reaching goals or be the driver of how, when and why things get done.
Comments