Advances in TV graphics
Jan 1, 2010 12:00 PM, By David Austerberry
New platforms leverage the power of the GPU.
TV2 Norway uses Vizrt applications to create and run election graphics.
Recent developments in television graphics platforms have transformed the “look” that can be created for a program. This is especially true for sports, news and weather. As the processing power available becomes greater yet lower-cost, graphic designers have been given more tools to realize their creative ideas.
Television graphics have come a long way since the closing titles were created using a caption roller in front of a camera. Lower thirds were just a black-and-white card placed on a stand in front of a camera. The black-and-white characters could be filled with a color matte on the production switcher, and an edge might be added to improve visibility — and that was it.
The invention of the electronic character generator (CG) offered graphic artists many possibilities for the lower third. It became possible to easily use a range of typefaces. Colored type and backgrounds could be set at will, and type could be bordered to aid visibility. A whole range of dynamics became possible that could not be created with the basic caption roller. The concept of the matte or linear key meant that graphics could sit on a semitransparent background, another aid to visibility.
The world of television graphics has been transformed since those early days, with all manner of 3D creations, virtual sets and data-driven templates.
24-hour German news station N24 creates interactive graphics to show the possible results of an election.
The creation of graphics is inherently labor-intensive, yet directors always want more to add that certain “look” to their program. Many of the advances in recent years have been to automate the on-screen graphics and reduce the need for the craft personnel in regular production.
Today, there are two main families of graphics products: One stems from the CG, the other from computer graphics and developments like the virtual studio. The CG has evolved as processing power increases year-on-year. The original products used custom hardware to create effects in real time. Computer graphics traditionally ran on (what was then high-powered) computer workstations. The creations had to be rendered by the CPUs to create the final graphic. For anything complex, the processes were much slower than real time, so to speed the processes, a render farm comprising many computers was used to shorten the run time of a job.
GPU advances
Changes in technology have made graphics platforms more affordable. Some video cards now include powerful graphics processors that can be used for 3D modeling and rendering. The huge market for 3D gaming means that these cards are manufactured in large numbers and offer good performance at a fraction of the price of a bespoke processor designed solely for broadcast graphics. Unlike the serial processes of the CPU, graphics processing units (GPU) use many programmable pixel shaders (thread processors) in parallel to perform tasks like transforms and shading, as well as offer an alternative to the render farm.
UK news provider ITN uses the Vizrt Engine to control video playback and as a telestrator.
A graphics platform today comprises a multicore PC workstation fitted with a high-performance graphics card and an HD-SDI interface card. Inherently lower cost than a proprietary platform, the modern graphics system can be deployed with a freedom that would have been limited in the past by caps on available capital and operational budgets.
The processing power available in a GPU-accelerated platform means that today 75 percent of graphics that were once rendered in advance of airing can now be generated in real time. And the CG has gone from the edit bay to graphics and titling applications running within the NLE.
The GPUs can be programmed using generic APIs, which free vendors from lock-in to a specific graphics card. One example is Open GL, a vendor-neutral, multiplatform graphics standard for developing 2D and 3D applications. NVIDIA has CUDA, a general-purpose parallel computing architecture that leverages the parallel compute engine in its GPUs. Using the GPU to solve many complex computations can create 3D graphics in a fraction of the time required on a CPU.
All of this available power can be leveraged by vendors to add features, or create whole new applications. For graphics designers, the power has provided new tools and new capabilities that can be used to realize the fruits of their imagination. For producers, the automation — templates and data-driven graphics — lowers costs without sacrificing the “look.”
Streamlining workflow
Last year much of the attention of broadcasters was to look for ways to improve the efficiency of their operation and to take cost out of the business. With graphics creation being a craft, and inherently labor-intensive, it presents a specific area of interest.
One solution is the wider use of data-driven templated graphics. The other is the management of the ordering of graphics from the art department and managing the fulfillment through to the news, sports or other departments. New business models, like operating graphics “in the cloud,” offer stations new ways to collaborate and to simplify basic graphics creation for the newsroom. Regional or group stations also look to maintain consistent branding across the organization, which is often a challenge with a geographically-dispersed business.
| Want to use this article? Click here for options! |  |
