Skip to main content

The advent of video servers changed the way we handle spots, news and long-form material. While this might be viewed as a revolutionary step for broadcast, we have since found ourselves in more of an evolutionary process.

The latest stages of this evolution are the growing importance of file-based workflows and the transition from SD to HD, which leads to the need for greater flexibility in media storage. The modern newsroom faces perhaps the most daunting challenges. This includes handling and storing a vast array of media types and their respective compression schemes, data rates and resolutions. This is typified by the ongoing SD-to-HD transition.

The approach taken to storage will not only dictate how much a broadcaster can store in any given storage system and how well it is protected, but also how many devices can be simultaneously serviced. Video newsroom production systems tend to go through pronounced cycles of high and low system use. It's important to design for the high-use times so that users or critical systems are not excluded. The way that storage is implemented as we move from SD to SD/HD workflows can have far-reaching effects.

News departments from local stations to major networks continue to learn about the impact on storage when transitioning to HD and how different storage technologies can be applied to solve their problems. For example, there are a number of scenarios and challenges related to the choice of codec and compression rate and how it affects capacity and bandwidth requirements. This equates to a variety of differences dependent on the broadcaster's technology choices.

The effects of media compression

There is certainly the possibility that storage capacity requirements will jump up dramatically upon transitioning to HD news. However, there are ways around this thanks to incremental video compression improvements that can provide good image quality at similar data rates.

One example is a station that migrates from DVCPRO50 (SD) to AVC-Intra (HD). By choosing AVC-Intra Class 50 (50Mb/s), the data rate for SD equals the data rate for HD. The number of storage hours remains the same, as does the performance of key workflow elements such as file transfer. Meanwhile, users with HD workflows can experience significant gains when switching from DVCPRO HD to AVC-Intra Class 50 — allowing their effective capacity and bandwidth to double.

Consider also the transition from IMX 50 (SD) to XDCAM HD422 (50Mb/s). The data rate is again the same for SD and HD, so the storage capacity and bandwidth are unchanged. Switching from IMX 50 to XDCAM HD (35Mb/s) or XDCAM EX will actually result in an increase in available bandwidth and storage.

Storage bandwidth

The transition to HD can translate to an increase in storage bandwidth requirements. Early adopters who chose DVCPRO HD found that they were significantly affected by the higher demands for bandwidth in HD when compared with SD. Server manufacturers either had to improve bandwidth within the server or introduce a completely different approach to the storage architecture. The latter amounted to the use of additional components and added a layer of complexity to the workflow. Over time, increased processing power within the server and high-capacity storage systems with much improved bandwidth and connectivity helped all but the largest network systems.

Ultimately, the codec and compression scheme selected dictates the bandwidth requirement. As newer compression schemes such as MPEG-4 H.264/AVC take hold, they will use less bandwidth and provide significant storage savings compared with MPEG-2.

Edit-in-place

The edit-in-place process gives editors access to news assets while ingesting. The ability for editors to work with content as it is recording or being transferred in as a file is crucial to quick turnaround and the ability to access news assets. Allowing for alternate sources such as P2, XDCAM or shared folders during this process further enables quick editing progress.

Editors can access the appropriate assets for a story directly from shared or attached storage, cut them into the appropriate sections and turn around the finished product to playout with expediency. It is an ideal example of how co-location of the record, edit and playout elements within a single storage system is pivotal to speedy turnaround.

The rise of citizen journalism has received its share of press, and citizen journalism applications available through devices such as the Apple iPhone successfully employ MPEG-4 compression. Handled correctly, the video assets can be intermingled with other codecs to create finished content for air that aids storytelling.

The many choices for shared or attached media mean that a system can be dealing with a wide range of codecs and associated data rates. A single timeline can be composed of HD content at 100Mb/s or higher and cell phone footage that is less than 1Mb/s. Determining peak usage bandwidth can become challenging, particularly if the high data rate content is used on an occasional basis. (See Figure 1 on page 26.)

Asset management

Newsrooms have an ever-expanding library of video files accessible at a moment's notice to insert into a breaking story. The video library is a strategic asset that must be preserved. The evolution of the library to include high-quality HD formats that can be used for HD or SD broadcast, mobile, webcast or on-demand IPTV is driving many storage upgrades.

News departments are increasingly looking to work with a single defined HD mezzanine format that will serve all their archive needs. (See Figure 2.) While older assets may be legacy SD or HD formats and wrappers, new material is standardized on high-quality compression and common wrappers such as MXF-OP1a. Overall quality is preserved with the constant need to decode and re-encode for different playout formats (HD to air; downconversion for SD and lower resolutions for mobile).

New architectures allow an editor from one timeline, with a single click, to create a final output and employ the NLE to create HD, SD and low-resolution browser (Web) versions — creating quick turnaround of multiple versions of the same content from the same storage system, all at first-generation quality. This first-generation quality feeds back into the archive as mezzanine quality for finished/aired media.

Continue to next page

New pressures and new solutions

There is increased momentum toward file-based transfer of content from the field, even as news editors receive field tapes delivered by hand and take in real-time feeds.

The push to transfer content into the storage system as quickly as possible doesn't differ whether received by tape or within a file. Still, the ability to work with a data stream as opposed to a baseband tape speeds the availability of content. File-based content sources can be imported much faster than real time, and all users have access to the content when the transfer begins.

The ability to move media with greater speed is enabled by prevailing IT technologies, such as 10GigE and PCI-Express. Catch servers and tapeless acquisition formats are able to take advantage of these changes. As a result, system bandwidth needs increase.

The result of these increases in performance is that editing, QC and playback systems can access content much more quickly. To enable systems that support these needs, the single storage system needs to be able to either support an increasing amount of bandwidth, or the components of the system need to be divided, normally by workflow operations, so that each step in the process is sustainable under peak use.

Workflow division across different storage environments that are linked by an overarching media management system not only allows for separation of raw ingest files, work-in-process files ready for final review and transmission-ready content, but also enables bandwidth management. (See Figure 3.) IT-based clustered storage technologies encompass expandable bandwidth architectures and also allow edit-in-place with user rights and associated folder structures. This ensures that certain users and groups can only access specific assets for specific jobs in their areas of responsibility, creating separate, compartmentalized environments for ingest, editing and playout.

This organization benefits the workflow process, and it allows the news organization to scale the system from as low as one-in, one-out plus editing, to hundreds of record and playout ports and editors. Storage capacity and bandwidth for SD and HD news content can strategically be addressed and scaled along the way.

Looking forward

Broadcasters continue to look for lower-cost solutions, and they often turn to IT-based nearline systems. These typically do not tackle all the needs for broadcast servers for shared storage.

Systems can end up with a less effective nearline store, with ingest and air being cached to smaller dedicated broadcast servers. This can end up being an operational advantage in scenarios such as large network systems, as it naturally segments the workflow and simplifies management of high media volumes.

Modern storage systems are able to accommodate the increasing need for bandwidth and storage capacity, from single, shared-storage systems that deliver on immediate access to all assets to different storage systems that divide the workflow. This gives broadcasters a number of viable options as they continue the move toward full HD systems.

Andrew Warman is group product marketing manager, servers, at Harris Broadcast Communications.