Over time, most technologies evolve to support the changing needs of the marketplace they serve. Television, in all of its incarnations, is no different. Nowhere is that more obvious than in the control room, where production switchers take center stage.
When digital effects first burst onto the scene with Vital Industries' Squeezoom around 1980 (the patent application was dated April 1979), production changed forever. Until then, it was not possible to manipulate the size of a frame, let alone the dynamic effects that Squeezoom, Ampex ADO and their progeny of today can do.
Imagine the revolution digital effects represented to the directors and technical directors of that era. Previously, the contents of a picture could only be composed of layered elements, principally keys and backgrounds. Digital technology, however, enabled the active picture to be resized and repositioned. Full 3-D manipulations could be done quickly, including page turns and mapping onto solid objects.
Today, we take these capabilities so much for granted that we fail to recognize just how revolutionary they were. Now if pictures are shot incorrectly framed, we simply reframe them. As late as the 1970s, however, that was unheard of. Digital effects were outboard devices that processed analog video and delivered it to analog production switchers as a key and fill. A four-channel DVE — an astounding piece of hardware at the time — filled a rack and cost in the mid-six figures, which is more than an entire switcher today. These units were considered so high-tech that the U.S. State Department would not allow the ADO to be exported to the communist block. Apparently, it was being used in military flight simulators.
Today, it's more than just a switcher
Contrast that complex situation with what we can expect today. Over the last 30 years, effects memory has evolved into complex control systems. Out of necessity, switcher manufacturers have incorporated sophisticated systems at the core of production switcher design. Some switchers feature a production automation application on the front end, controlling outboard video servers, character generators and even remote controllable cameras.
The power in modern switchers includes still stores and clip stores, color correction and, in some cases, aspect ratio conversion, upconversion and downconversion. Until recently, switchers were designed for a single production standard, with software configuration possible for support for 525/30 and 625/25 standards.
With the introduction of HDTV looming, manufacturers designed a generation of switchers with sufficient memory and flexible I/O ports to allow conversion of the hardware from SD to HD. This allows a switcher purchased today to have utility into the future, which of course makes the finance whiz in a broadcast facility easier to tame. With longer product life cycles, it is easier to achieve an acceptable ROI.
Accommodating SD and HD
In the last three years, it has become increasingly important for production systems to accommodate both SD and HD material at the same time. There are three main strategies for making that happen.
In the first, content can be processed in two parallel chains, with SD and HD segregated into systems intended for only one format. This is obviously complex and expensive. In principal, it allows the best application of graphics elements that don't need to be unisex, or equally appropriate for both 16:9 and 4:3 frames. One switcher panel may control two electronics frames as long as inputs are carefully mapped to each frame.
A second approach involves converting all content to one format, usually HD in the interest of improving quality, with the SD picture being derived from the HD picture after production switching. This is quite appealing from a cost standpoint, but choices must be made on how to handle graphics to protect both frames. Engineers must also consider what happens to the aspect ratio of SD content in the process, so the output media will be appropriate to intent in both aspect ratios.
Such an approach may lead to graphics compromises that are equally inappropriate for both release formats. It is, however, easy to understand, and upconverter manufacturers love this approach because it sells products.
A third approach takes advantage of the ability of production switcher systems to accomplish upconversion as part of normal video processing. In some cases, this amounts to providing tie lines to external converters so that any incorrect source is converted before use.
In this manner, it is similar to the second approach. If the switcher has converter capability embedded within it, however, input flexibility does not hinder system design or burden the approach with excessive hardware for the sake of ease of use.
Some hardware will do the required conversions and, as a bonus, can accept essentially any common SD or HD standard at the input. In addition, it can output multiple standards directly from the switcher electronics. This approach provides the most flexibility for production staff because they don't have to spend extra time planning to make sure they can handle the input and output formats.
If one does a cost analysis on this method, it is easy to see how dollars can be equivalent or lower than an approach with multiple format converters. This also simplifies latency planning, leading to less complex audio delay matching issues.
You may be asking what this has to do with the Squeezoom. All digital processing of picture content first showed up in digital video effects. That grew in capability to allow hardware processes that are now tightly integrated into production switchers.
So what does this mean for the future? We are approaching an era when multiple processors in a blade server will have the raw processing power needed to replace special-purpose video processing engines in video production switchers. When that happens, there will be some stunning advancements in image processing.
John Luff is a broadcast technology consultant.
Send questions and comments to:firstname.lastname@example.org