25 January 2012

Broadcast to IT migration driver: Bandwidth

The adaptation and adoption of IT systems for broadcast use is governed by a number of motivations; in this series of posts I hope to talk about (nearly) all of them.

The first is bandwidth. In broadcasting data streams often have fixed bandwidth. For example, HD-SDI has a bandwidth of 1.485 Gbit/s, if the encoding process doesn't actually need all the bits then it just fills to stream with filler data, which serves no purpose except to keep the data stream fixed. This is obviously wasted bandwidth, so in some systems (especially in IT) there is a variable bandwidth. In fact, when a PC is turned off, the cables that connect it to a router are pretty much empty.

In broadcasting a variable bandwidth can be used. For example, when applying compression, a frame showing a plain wall will require less data to encode that a frame showing a complex flower bed. So the spare bandwidth could be used for something else, like sending meta data.

Middleware, and why broadcasters need it

Web 2.0 is often chiefly defined by it's 'interoperability' - which is the ability to work conjunctively with other technologies. Middleware is the software that makes this happen, a common example could be an API (application programming language) which are used for things like social networking plug-ins; the provider (e.g. Facebook) publishes an API for it's services, and this is then used by the developer (e.g. Wordpress) to link directly to the service. Because this middleware is designed for any potential connection to the service it can be fairly generic, reducing the need for the developer to directly contact the provider.

Because middleware allows interoperability, it is an important part of broadcast IT systems. Any IT system replacing a particular segment in broadcast technology must be able to match or improve it's interoperability. This is because assets in broadcasting are often managed by several stakeholders which may have differing equipment. However, broadcast standards (such as SDI) are standardised by organisations, and then often enforced by licencing.

Issues arise when IT systems convert standards into file systems, at which point they stop conforming to a specific standard and cannot be read by equipment designed to read that particular standard. As a result, the IT systems must have appropriate middleware to allow different stakeholders to manage assets in IT systems.

Data transfer languages such as XML can be very useful in creating such middleware.

Source: Own knowledge.

Magnetic degradation of hard disk drives

Archival scientists strive to develop technology that can store media forever with out loss to the original content. This desire is evidentially justified by broadcasters, who often create historical docs using old footage, or who wish to sell on material to other broadcasters. Many archival technologies (such as tape) can decay magnetically.


In hard disk drives (HDDs), magnetic decay is usually a result of heat build up over time which increases sputtering* and this natural de-magnetisation.  Although for data to corrupt through this process is rare and could take decades in a modern hard drive, it much more likely that the hard drive would outright fail before time, which prompts the need for a RAID system.


One reason modern HDDs do not de-magnetise very quickly is due to a clever process called giant magnetoresistance (GMR) which was discovered in 1988.


*Sputtering is  is a process whereby atoms are ejected from a solid target material due to bombardment of the target by energetic particles. (http://en.wikipedia.org/wiki/Sputtering)



Sources:
http://en.wikipedia.org/wiki/Giant_magnetoresistance
http://www.smarthdd.com/en/bad_block.htm
http://bytes.com/topic/software-development/answers/851490-magnetic-decay-modern-hard-drives