Production switchers

Digital production switchers have been on the market for many years now. The first digital model was the Grass Valley Group Kadenza, introduced in 1988. It was a unique architecture intended for the postproduction market and pretty much unsuitable for live production. It had some operational modes in which it could be driven like a more conventional switcher and others in which each successive row of crosspoints acted like another layer in a complex mix. Up to five layers could be controlled at one time, with a total of eight layers available. The heart of its technology came from the Kaleidoscope digital effects generator. Each picture layer could have full DVE capability, or simple “switcher channels” without transform capabilities that could be used.

The Kadenza was equipped with parallel digital component inputs. Cable lengths had to be kept to under 100 feet. Building cables for digital component equipment was a bit of an art, with tight specifications on the electrical length for each pair carrying one bit of the data stream. By using analog-to-digital converters, one could use NTSC and component analog signals as sources or destinations.

While the Kadenza was a groundbreaking product, it was quickly joined by competition from Alpha Image and Abekas. The Abekas A-84 had eight fully functional layers, while the Alpha Image catered to a smaller market with four layers and nine inputs.

For a while, as standards organizations and manufacturers debated, composite and component standards both existed, as did both variations of production switchers. Sony even made switchers (DVS-8000) that could be reconfigured quickly between both standards. However, as equipment designed for the SMPTE 259M standard became less expensive and more ubiquitous, component strategies prevailed.

Today, component equipment is the basis of virtually all new systems, and the old parallel interconnection standard has disappeared. Though many digital switchers can still be equipped with analog inputs and outputs, virtually all production switchers delivered today process in the digital domain.

There are good reasons for this beyond the desire of manufacturers to sell products to the marketplace. Digital switchers are inherently more reliable and require far less routine maintenance. With no, or very few, analog circuits, there are no adjustments necessary to keep performance optimized. Manufacturers have been able to deliver features that far exceed the capability of analog switchers for real dollars that approximate the cost of the best of the last generation of analog products. In the last year a new crop of less expensive digital switchers has emerged that provide the benefits of digital technology and reliability to a market more sensitive to capital cost.

Digital switchers also lend themselves to several interesting possibilities. With Kadenza, GVG was first to introduce a powerful function in which the “assets” of a digital production system were networked together. In a facility that had two control rooms it was possible to give control of some layers to a second room, shifting capabilities to where they could be most productive without duplicating all options in each room. Abekas expanded this function with the A-82/83 series, allowing entire effects banks to be assigned to a second control panel.

Some functions that existed in analog products have been significantly enhanced with digital technology. Timeline control over effects memory existed in a number of analog products. However, digital switchers require less technology to “memorize” settings and transition between them. Such features can also be modified and enhanced to best suit the marketplace without new hardware development. By forwarding new operating software to users, manufacturers can provide updates and new features to the entire installed base. Doing so has the added benefit of keeping the support costs lower since fewer versions need to be fully supported at any one time.

Modern digital production switchers now include full function digital effects in software (and hardware) within the switcher. No external connections are necessary and the DVE channels are tightly linked to the switcher control system. At first, the DVE channels supplied in such implementations were simple planar devices with limited creative capability. Now most implementations have the ability to at least supplement external DVEs, if not replace them.

Today it is quite common for links between newsroom computer systems and other devices, including production switchers, to be a central part of any production. By allowing the script and all production elements to be tied directly together, one creates a seamless production environment, cutting down on the second-by-second management of a complex program. Digital switchers greatly enhance this capability, but not without some new problems. There is currently no universal language for such devices to use for communication. This means that each pair of devices must be “debugged” individually, which is not always simple to coordinate since manufacturers of many of these products compete with each other.

One manufacturer has carried this concept to a logical extreme, integrating newsroom automation, production automation, a digital production switcher, graphics and character generators, and camera remote control into one seamless product. Developing the script for a program in this system would in theory allow a program to be controlled by a single person. The potential drawbacks are clear: This is the ultimate all eggs in one basket. However, if the basket is reliable and allows for creative freedom to handle live events, it could help control costs in some production environments.

Production switchers have also grown in scale. The conventional analog switcher was 24, or perhaps, 32 inputs. Today, one manufacturer is delivering a switcher with 80 physical inputs and a virtual router internally that can map any of those inputs to an active effect — without requiring the input be assigned to a physical button! Even the outputs are programmable, essentially making the backplane of the switcher virtual as well.

In the last year we have seen the emergence of the next serious step up in production switcher capabilities. Until now a company who desired to have future HDTV capability had to assume its current 525 switcher would have to be replaced when it moved to HDTV. However, at least two companies now are planning, or delivering, switchers that can accept both 270Mb and 1.485Gb inputs. You cannot mix signal formats (yet), but at least you can shoot today in 525 and tomorrow in 1080i, 1080p24 or 720p. This provides a significant insurance policy for the future.

John Luff is vice president of business development for AZCAR.