Real-time demands and possibilities

A few days ago I accompanied my children while they watched an episode of El chavo del ocho made around 1973. And I noticed with some nostalgia that the end credits were a sequence of rolling titles that looked definitely electronic but made with a strip of paper. When mounting the end of the program, a firm pulse assistant had to move a crank to get the titles to slide smoothly on the screen.

In ten or fifteen years we went from Chinese ink to electronic graphics. From the expensive workstations of the eighties we move on to the world of personal computers and graphics production based entirely on software. Therefore, while trying to generate static graphics and animation loops , it is perfectly possible that the graphic package of a cable channel can reach the same degree of quality as the pieces made by the most powerful television network. In reality, the problem is more one of talent than technical resources.

But there is one front on which the differences are still being felt: the automated production of real-time charts. We are talking about animated sequences that are updated by querying remote databases. Of virtual scenarios with unlimited camera movements. Of complex animated headlines that include images of the news of the day, every day.

The truth is that with this type of resources, the production of graphics for television is reaching a new level of quality and generating new expectations in an audience that, for better or worse, has learned to enjoy the look of the most sophisticated graphics.

Those of us in the industry for a long time remain convinced that light, framing and sound are the most important production values. But in the world of zapping, graphics are the factor that makes it possible to distinguish one product from another. The look of the graphics is the brand, the particular packaging of each product or each channel. And the bitter truth is that some mediocre products manage to compete in favorable conditions thanks to good branding and attractive graphics.

Vizrt is a company that has concentrated on the creation of highly productive tools for the creation and distribution of graphic pieces in real time or very close to real time. The firm appeared in mid-2000, from the merger of RT-Set and Peak Broadcast Systems.

We talk about the union of two powers: RT-Set was one of the companies of Israeli origin that left the market of military simulations to help invent the technology of virtual scenarios, and Peak was the first company that offered automated graphics systems for television in addition to being the main heir of the companies. As who invented automated TV in the eighties.

Vizrt created a fully modular graphics production system, which allows to integrate appropriate solutions for any application from the creation of 3D animation on a desktop computer, to titling systems for newsrooms, weather graphics, virtual scenarios or interactive pieces for the web.

The heart of the system is Viz| Engine, a real-time rendering system that works collaboratively with other design and graphics development applications, both vizrt and the most popular pieces of software on the market. More than a graphics engine or a compact render farm , Viz| Engine is a playback system that manages video streams, task lists, and metadata collections.

Current versions of Viz| Engines can generate SD signals using Pinnacle Systems' Targa 3200 card and HD signals using the Panasonic AV-CPG500 video processor, released at this year's NAB. Although for SD Viz applications| Engine can operate on ordinary PCs, the bandwidth requirements of the 1080i signals generated by the Panasonic processor make it necessary to use RISC stations, from Silicon Graphics.

For several years Viz| Engine uses OpenGL technology to generate 3D graphics, which are then copied to separate devices to generate the video signals. However, a few days ago the company announced a strategic alliance with Nvidia to use its QuadroFX accelerator cards in the assembly of the new Viz stations| Engine.

This innovation will represent a significant improvement in Viz's performance| Engine, because the new Nvidia cards generate SDI signals with transmission quality, which will make it possible to do without additional pieces of hardware; By generating graphics and signals on the same card, it is possible to overcome the limitations of PC architecture and offer performance comparable to that of expensive high-end graphics stations.

In fact, at IBC 2004 Vizrt presented a PC-type workstation, equipped with an Nvidia QuadroFX 4000 card, generating up to seventy layers of graphics in real time which represents enough bandwidth to handle HD signals on PCs in the very near future. It will even be possible to handle several graphics channels on a single workstation.

By assigning the creation of signals to a separate station, Vizrt manages to simplify the development work, which can be advanced in workstations that could be qualified as normal within the high-end of the PC world.

Vizrt offers software packages suitable for any need, and constantly announces its interest in partnering with operations that need to develop special applications. The development stations function as satellites of Viz| Engine, which use commonly used solutions for storage and connectivity.

An interesting example is Viz| Weather, the module for creating climate charts. It is a package that can operate on a fairly normal computer and that directly handles connections with satellite services; it can be integrated into automated newsrooms of any provider, can interact with conventional character generators and manage a wide repertoire of templates that allow you to change the look of the product in a minimum time.

Viz| Trio is a character generator. Viz| Ticker, a manager of textual, financial, sports or any other type of information. Viz| Maps can generate maps of any region of the world from textual requests or GPS data.

Viz| Arena is a special case. It is a specialized system for inserting graphics in sporting events, which goes far beyond the simple overlapping of lines and arrows. Arena, when tracking objects, enables applications such as a credit that chases a character, a flash around the ball, a flash around a foul or a marker that allows to locate the elusive puck in a hockey game.

On the other hand, Viz| Virtual Set follows in the tradition of high-profile virtual scenography systems, which generate real-time graphics from tracking information from cameras and lenses. However, the simple brute force of the most modern hardware makes it possible to offer unlimited camera positions: the polygon count of Viz systems| Engine makes it easy to deliver on this promise.

In addition to specialized graphic content generation systems, Vizrt offers high-performance solutions for the production of graphic elements, 3D animation and visual effects of all kinds. The center of this proposal is Viz| Artist, a painting and composition system that integrates transparently with the most popular graphic applications on the market. In addition, Artist supports a fully open plug-in architecture and, as you might expect, seamlessly integrates with modules dedicated to specific tasks. Although some users claim that Viz| Artist has a fairly steep learning curve, you have to admit that once you learn to manage it, it is incredibly productive.

Vizrt's current offering is rounded off with the modules needed to mount fully automated, graphics-intensive transmission operations. Viz| Media Sequencer and Viz| ContentPilot is a powerful content management solution, which allows you to implement any programming model that is possible to imagine for television or any other medium. In fact, the success of the model proposed by Vizrt lies in its flexibility, the synergy between the modules. Let's pause to present an example: the coverage of an election, an event that can intimidate any group of engineers.

Viz| Ticker can connect to a SQL database and capture the results of official bulletins. After previewing with a text crawl , Ticker formats the text to send to Viz| Trio and compose 3D charts and graphs, according to templates previously defined in Viz| Artist. Trio generates alarms when there are significant variations in the data and fires an independent animation as a news flash, but simultaneously triggers a process in Viz| ContentPilot that generates a pop-up with the news on the channel's website and makes notifications of the case to all reporters connected to a Sony digital newsroom .

Finally, Viz| MediaSequencer is used to generate video streams with the data that constantly feeds the simulated video walls of the virtual set , specially designed for the occasion. And if there is a storm and you need to interrupt the broadcast for a weather bulletin, the response time is practically zero. All this assembly can be operated by one or two people only, and uses a single Viz station| Engine.

What lesson does Vizrt leave us with this perfectly achievable example? We have to start thinking about our business based on workflows. Productivity is no longer in having fashion toys at your fingertips, but in knowing how to combine the available tools.

Vizrt offers us some clues about an immediate future, in which media operations could finally cease to depend directly on human operators. Organizations as important as CNN, most European state television stations and the main news operations in Southeast Asia have already opted for this model. What do you think?