It’s hard to imagine today, but it was really not that long ago that switching between video sources in a live production was an engineering task, primarily concerned with keeping signals in synchronization and minimising distracting visual jumps.
It was not until 1969 that the first device we would recognize as a ‘production switcher’ was launched into the market. The Grass Valley model 1400 put the process of ‘switching’ sources, and introducing simple transitions between them, into a simple and logical controller, allowing the operator to focus on the creative flow of the program without worrying about the underlying technology.
The basic layout and functionality of a production switcher was established. Each video source is available under two rows of buttons, for preview and on air. You can cut between sources simply by pressing the appropriate button to take the new source, or you can make a transition between the sources, controlling the speed of the wipe, mix, dissolve, or effect using a T-bar.
As so frequently happens, operators loved the simplicity and creativity of the basic production switcher – and almost immediately wanted it to do more.
Keying was an early addition, so that one signal could be added on top of another. At first, this was to superimpose a live bluescreen shot onto a background, or a graphic onto a scene. This meant that more sources were required, so there was pressure to build larger switchers. An operator could build a static graphic and leave it on a switcher input to be keyed onto the output as required. But again, creativity pushed the technology and directors asked if the graphics could move. And maybe they could be linked to some other effect, so that a graphic sequence could cover a wipe.
Of course, technically, there is no difference between keying a moving signal and keying a fixed graphic. The challenge comes during the switch, when the transition and the foreground need to be synchronized with the right point in the background. Shouting “go” or “cue” might be good enough, but it adds to the chaos of live production and is open to human error.
Live production has to be crisp and precise, so production switchers evolved to be able to control external events and devices. Once set up, pressing a single button set off multiple events. With this ability, the opportunity was taken to include control of a router, so many more sources could be included in the production than there are buttons on the switcher.
Many sport productions call for a very large number of sources. The 2010 Fifa World Cup will have at least 30 cameras covering each game, plus replays, graphics, effects, and more. The major golf tournaments may have two or three cameras on each hole, plus roving wireless cameras, running up a very large number of inputs.
Today some of this functionality is included inside the production switcher. Clip and still stores, for instance, are frequently internal, which means that regularly used sequences and stings are instantly available, usually from solid-state memory.
Digital video effects (DVE) were also originally generated in standalone chassis, but are now always included in the switcher. At that time, the largest production switcher had six channels of internal DVE, which was considered excellent. Today’s models have 20.
Bringing additional functionality into the switcher is more efficient in operation, but it also tends to save costs by requiring fewer discrete pieces of equipment, less rack space and so on.
Users were once again happy, but for a relatively short period of time before pushing the envelope still further. They loved the excitement of a layered effect, which included a moving key over a transition between live sources. They loved it so much, directors wanted to do it many times during a single live production.
Setting each layered effect called for a lot of actions on the part of the operator, to set the right paths, cue the appropriate clip and put it on the keyer, select the effects and more, all so it happened when you pressed the button or moved the T-bar. Under the pressure of a live program, there was a lot that could go wrong.
The logical step was to allow an effects sequence to be programmed in advance, stored, and recalled when required. Grass Valley gave this functionality the name ‘effects memory’ or E-MEM. Originally, switchers had 100 E-MEM registers for storing various effects: the latest models have 1000.
Certainly productions have become more complex, but it is very rare for a production to actually require 1000 different effects sequences. The number allows the operator to set them up in a logical sequence, speeding recall. Having more storage slots also allows effects for multiple shows to be stored together, or effects for different operators.
E-MEMs have a timeline: they describe how a complete effect is going to appear on screen. To take a very simple example, you might want to add a graphic to the program output, mixing it in over 10 frames. You could, of course, do this – over approximately 10 frames – by setting up the graphic in one of the keyers then using a T-bar to bring it in. But doing it with the press of one button is much more secure.
You can make them more complex than that, and a single E-MEM can have a number of keyframes marking different parts of the transition. For example, a broadcaster might want to come out of a replay sequence back to the live action with a DVE move that first tilts the replay source to bank up and away, then adds a caption on a virtual 3D border at the bottom of that plane, then completes the move away leaving the live shot, and finally adds the clock and score back on.
That sort of sequence will be a signature part of that broadcaster’s coverage, designed once and expected to be repeated on every program. So E-MEMs, once created, can be written to a memory device – probably a USB memory stick today – and loaded into the switcher during set-up.
Anyone who watches live television today, and particularly sport, will have noticed that there is a lot going on. If you watch it with a careful, analytical eye you will realize that the production switcher needs to be putting effect on effect, with each layer having its own set of graphics. The simple model of production switcher will no longer suffice.
In effect, it is repeated a number of times, adding flexibility. Each of these switchers-within-a-switcher is called a “mix and effects bank”, usually shortened to M/E. An M/E bank is a complete production switcher with keyers, DVE, transitions and the full set of input sources.
Then you switch, mix, or DVE transition between M/E banks to create the final output. This is how, in the example I suggested earlier, you could have a complex event (the replay sequence) being moved out and replaced by another complex output, through a complex transition.
Adding more M/E banks adds to the creative power of the switcher, although it inevitably adds to its physical size, too, and after a while the operator’s arms become a limiting factor. One solution to increasing flexibility without increasing the size is split M/E mode, which Grass Valley calls Double Take.
Because each M/E bus has many keyers available, along with a number of utility background buses, you can produce similar but distinct outputs from each M/E bank. The most obvious simple use is if you were covering a football match to be shown in both the home country and that of the visitors, you would want the same pictures but different graphics, perhaps with the branding of each broadcaster or the captions in different languages. Another application of Double Take would be in the production of stereoscopic 3D television, where the two “eyes” would be two channels going through a single split M/E, so the same cuts or mixes happen to them at the same instant, graphics could be positioned in 3D space, and so on.
This is just a snapshot of the way that the production switcher has evolved, in 40 years, from a simple means of cutting or mixing between sources to, a massively powerful creative resource.
The most important point to make about this journey is that it has been driven, every step of the way, by program makers always wanting to do something more, to create television which is more exciting, more engaging, more immersive. I have no doubt that this is not the end of the line: products will continue to evolve, and continue to satisfy the needs of the creative community.