The role of digital asset management systems in broadcast organizations has been a hot topic in recent years, and an increasing number of these systems are being implemented. A primary driver for this trend has been the shift to file-based media, which makes collaborative and “searchable” workflows possible; other factors are the growing pace of media production and the economic drive to improve efficiencies and cut costs. (See Figure 1.) As a result of these changes, media creation and archiving now have the potential to look more like private clouds of data, with widespread accessibility across the workforce, whereas in years past, video assets were not easily available to all who could make use of them.
Figure 1. Media asset management systems are increasingly being implemented by organizations due to the accessibility and ability to search content across all storage, from any device (desktop, laptop and mobile).
Asset management solutions
Broadly speaking, asset management can be categorized as either digital asset management (DAM), a term that covers enterprise asset management across a wide range of media types, including images and documents; media asset management (MAM), which covers systems that emphasize video and audio capabilities; and production asset management (PAM), a subset of MAM systems that targets work in progress rather than long-term archival capabilities. A wide range of solutions are available, with pricing ranging from close to $1000, up through enterprise systems costing close to $1 million. The systems at the low end of the scale tend to emphasize out-of-the-box capabilities and speed of installation, while the larger deployments often involve a high degree of customization for a specific site’s requirements.
Archive storage options
Turning to storing digital video assets, there are special requirements when compared with the data storage needs of a typical IT department. The main differences are the much larger total volumes of data and larger file sizes. Often there is also a need to consider storage of video assets on a much longer timescale than is relevant for general IT. However, an important shared requirement is to have strong data protection, including a mechanism to recover assets in case of fire or other major disaster.
The main candidates for storing large volumes of video assets are disk drives, Linear Tape Open (LTO) cartridges and cloud storage, and there is a place for all three.
For work in progress, disk drives, typically in a RAID configuration, make a lot of sense, but for long-term storage of video assets the total cost of disk-only solutions becomes prohibitive when issues such as the cost of electricity, the cost of replacing disks and the typical need to have replica data in two different locations to provide adequate data protection are factored in.
Combining disk and LTO with intelligent management software provides an attractive option for storing large volumes of high-resolution video assets for the long term. LTO cartridges have a 30-year lifetime in normal office conditions, provide high data transfer rates and a low cost per TB. Today’s LTO-6 cartridges cost about $100 each and provide a native capacity of 2.5TB, which means that a single LTO cartridge will hold more than 100 hours of HD recorded at 50Mb/s or more than 50 hours at 100Mb/s. At about $40 per TB, the cost to produce a duplicate LTO cartridge for offsite retention is attractive.
Cloud storage is a service, and the service provider uses either disk or LTO to store the digital assets. It naturally provides global access to the assets, and the service model can reduce capital costs. Today’s cost of bandwidth means that it is expensive for high-resolution assets but attractive for low-resolution proxies and metadata.
Implementation challenges
Across the board, there are a number of key challenges that tend to recur when implementing an asset management solution. Among these are: 1) the need to accommodate legacy assets; 2) the need to fit with the existing infrastructure; and 3) the need to adapt to (or change) existing workflow expectations.
We will now discuss each of these in turn, and try to summarize best practices for handling the challenges.
Managing legacy non-file assets
There is a wide range of approaches for managing legacy non-file-based assets, such as those held on videotapes. These range from a simple switchover, with all assets after a certain date being file-based, to converting non-file assets on an as-needed basis, through to a major conversion project, turning all historical assets into files.
Figure 2. The preferred approach in implementing a MAM is to use a file-folder structure, such as the one shown above, accessible via a standard network protocol — rather than a proprietary interface between the storage repositories and the asset management system.
Most broadcasters and content producers will benefit from a simple cost-benefit analysis to this range of approaches. In many cases, a cutoff date combined with as-needed conversion of specific earlier assets will provide the best solution. However, there are clear examples, particularly in the case of organizations with large archives and rising costs and risks associated with potential data loss, where the approach of a major conversion makes the most sense.
For a major conversion to file-based assets, the process should be streamlined as much as possible to minimize costs. Where possible, the high-resolution assets should be stored on the final archive storage media, such as disk or LTO, in the required format as early as possible in the migration process to avoid unnecessary further processing by the asset management system.
Supporting existing and future infrastructures
The need to fit with existing production infrastructure is often a key determinant of what system(s) will be chosen. A “forklift upgrade,” in which editing systems, storage hardware and networks are all upgraded, is rarely ideal, especially in these times when capital expenditures are under pressure. Most MAM and PAM software vendors acknowledge this, and offer ways to implement their systems in the context of hardware and software already in place.
Where possible, customers should seek to follow industry standards. A key consideration for the archive storage, whether disk, LTO or cloud-based, is the interface to the rest of the infrastructure. This is important because within the lifetime of the video assets, often measured in decades, there will likely be a need to change the asset management system or even migrate the entire contents of the archive.
The preferred approach is to use a file-folder structure accessible via a standard network protocol such as CIFS or FTP rather than a proprietary interface between the storage repositories and the asset management system. This allows access to the asset repository which is independent of the asset management system and maximizes the options for future upgrades.
User acceptance
Finally, as in so many change management scenarios, the biggest hurdle is often the people who will be asked to use and administer a media asset management system. Generally speaking, they will have developed their own ad-hoc workflows without the help of an asset management framework. Most often, these are based on standard folder structures and internal best practices for folder organization, rights and privileges, etc.
Some MAM and PAM vendors provide approaches that attempt to streamline the transition from a folder-based workflow to a MAM/PAM. But perhaps the most effective way to handle this transition is to choose a system that allows the users to maintain their familiar views of, and access to, the assets being managed. When this can be achieved, the transition costs to MAM or PAM from folder-based workflows can be minimized.
Just as important as the cost savings are the increased likelihood of compliance. Although MAM and PAM systems can have a utopian appeal to management, the grassroots success or failure of such systems is often determined by the simple fact of whether users choose to “buy into” the system. If they ignore it, the capital investment — whether small or large — will mostly be wasted.
—Sam Bogoch is CEO of Axle Video, and Dr. Philip Storey is CEO of XenData.
