HD broadcasts over DTV channels hold the promise of delivering vastly improved pictures and sound, along with data that can augment the linear video medium in an age of hypermedia.
The HDNet network broadcasts all of its content 16 hours a day, seven days a week in the 1080i high-definition format. Shown here is the interior of one of HDNet’s remote trucks in action at the Olympics in Salt Lake City. Photo courtesy HDNet.
Ernie Kovacs once said, “Television is often called a medium because it's so rarely well done.” Perhaps HDTV programming will change this cynical view. Perhaps the creative community's vision combined with the sensitive application of technology can yield innovation and high-quality programming.
This article discusses some of the issues involved in bringing big-screen values to the small screen. It begins with the politics and technology of HDTV content delivery, and then focuses on HDTV production, including the processes necessary to develop, acquire, edit and distribute HDTV content.
On April 4 of this year, FCC Chairman Michael Powell outlined a plan to speed the DTV transition. Among other things, the plan seeks to advance two key goals: (1) increasing the level of compelling digital content available to consumers; and (2) providing cable subscribers access to that content over their cable systems. He proposed voluntary action to produce and deliver content that only DTV can provide.
His chief proposal calls for ABC, CBS, FOX and NBC, along with HBO and Showtime, to provide high-definition or other value-added DTV programming during at least 50 percent of their prime-time schedule, beginning with the 2002-03 season. Value-added DTV programming could be high-definition, innovative multicasting or interactive — as long as it gives consumers something significantly different than what they currently receive in analog broadcasts. This would require something more than a single stream of standard-definition digital programming.
Cable, DBS, equipment manufacturers and broadcast licensees each received their own admonishment from Chairman Powell to hold up other parts of the consumer-delivery chain. It is important to note that much of the prime-time schedule of the four broadcast networks is already available in HD, in some cases already exceeding the requested voluntary action. By allowing broadcasters to provide content that is “significantly different” as part of his plan, Powell tacitly admitted that broadcasters have significant reservations about HDTV production. The production community and the consumer electronics industry are not ready to deliver on the full promise of interactivity and other enhancements DTV can provide. What is left in Powell's list is multicasting and HDTV.
Figure 1. Some new HD cameras and camcorders can “overcrank” the frame rate for slow-motion effects and “undercrank” it for high-speed effects. This diagram shows examples of overcranking and undercranking for a 24p HD format. Image courtesy Panasonic.
But, if the production side of the industry is to increase HDTV content, it will have to gear up significantly. Four years ago, only a small number of facilities were capable of HDTV production. Since then, the number of commercial mobile-production units widely available for HDTV productions has grown. And other facilities have joined them, many in advance of significant consumer demand and at considerable cost and risk.
Over the last five years, the technology for HDTV production has improved and its cost has plummeted. Whole camcorders are now available for less than the cost of a single lens a few years ago. Cameras no longer need to be tethered to large VTRs that require significant power and space. Indeed, for little more than the cost of a high-end 525 EFP camera of only a few years ago, it is now possible to have an HDTV camcorder with equivalent features and reasonable power consumption. Hitachi, Ikegami, JVC, Panasonic and Sony, among others, have developed a range of camera products that offer features that, in some cases, were not practically available to the production community in 525 and 625 products. The sensitivity of HDTV cameras rivals that of their standard-definition counterparts. Colorimetry has been improved, especially in low-light areas of scenes. And some cameras can now mimic the ability of film cameras to “overcrank” and “undercrank.” These improvements in camera technology and production effects will materially affect the transition from film acquisition to video acquisition, at least for television release.
This year saw a commercially successful use of HD cameras for producing a major motion picture. Also, Lucas Film, Sony and Panavision have pioneered significant developments that production professionals will need to produce HD releases in the future. Clearly much of that development work was directed at the unique needs of release to theatres, but much is directly applicable to television as well. Chief among the developments is the use of 1080p24 acquisition. Using a 24p high-resolution master will allow conversion essentially to any release format — in much the same way that the film community is accustomed to working.
Figure 2. This box illustrates the framing compromises that images acquired in the 16:9 aspect ratio suffer when they are downsampled to 4:3.
At the same time, recording technology has also improved. HD recorders of a decade ago cost hundreds of thousands of dollars. They used open-reel tape that cost over $1000 per hour and did not last as long as tape used for conventional 525 production. Today, Panasonic and Sony offer cassette-based HD recorders with performance that was unavailable a few years ago. Both lightly compressed and moderately compressed HD images can be captured and used for post-production applications.
Terminal equipment that provides virtually all of the capabilities that standard-definition systems commonly use is available from many manufacturers. Converters between analog and digital formats (all HDTV systems use SMPTE 292M for interconnection at 1.485 Gbits/s), frame syncs, DAs, audio embedding and dis-embedding, and optical links are all readily available from manufacturers such as Evertz, Miranda, Leitch, Thomson, Sony, Snell & Wilcox, AJA Video, Cobalt Digital, ADC/Nvision and others. Where once the range of products was limited, it is now quite broad and supported in standard interfaces from a number of manufacturers.
Production and routing switchers have also progressed significantly. The first HDTV production switchers were extremely expensive. Newer models are still more expensive than their standard-definition counterparts, but they have come down quite a bit in price. Generally, they support the range of normal production effects in HDTV products to about the same level as they do for standard definition. The state of the art for routing switchers is much the same, with newer models erasing much of the price penalty associated with wideband products. Many are well integrated into the product lines to which they belong, which facilitates multiformat facilities with few penalties.
In short, broadcasters wishing to ramp up HDTV production face few hurdles to prevent them from using techniques that replicate those used for standard-definition episodic and live programming. The complications come from new techniques stemming from the very nature of HDTV vision and sound.
The nature of the beast
HDTV is all about three things: wider screens, higher spatial and temporal resolution, and improved digital (and surround sound) audio. Each carries with it additional knowledge that the production professional must be aware of when preparing to acquire and edit in HDTV. Higher resolution puts a burden on the director of photography to effectively deal with the higher modulation-transfer performance of modern HDTV images. Cameramen familiar with shooting at a given depth of focus in standard definition will find that HDTV cameras are sharper and may require rethinking production techniques (lighting, camera angle and/or distance from the scene) to achieve the desired results. The same is true when comparing 8mm film production to 16mm, 35mm and 70mm media. HDTV cameras seem to be more critical on focus. They certainly show the effects of poor focus more quickly. Sports camera operators, for instance, find that they need to practice more care in following focus as the primary object in a scene moves within the frame.
HDTV offers a choice of temporal resolution between interlace and progressive formats, and several choices of spatial resolution. The most popular production formats are 720×1280(60p) and 1080×1920(30i, 24p). The majority of production is done in 1080i30, with a growing amount done in 1080p24. More than one major distributor has standardized on 720p60 for production and distribution, despite being limited to a smaller circle of manufacturers that fully support the required production equipment.
But 720p hardware capable of variable frame rate (1 to 60 fps) is available, and is particularly attractive for certain types of productions. Manufacturers have used a clever technique to achieve this without changing the recording medium in ways that would make it incompatible. The camera captures the specified number of frames per second, but the recorder continues to record 60 fps, repeating frames as required to “pad” or fill the sequence. The frames needed for post production are “flagged” and processed during post production (see Figure 1).
Of particular concern for HDTV production is the wider aspect ratio. A production that must release in 4:3 (1.33:1) and in 16:9 (1.77:1) has to be shot without compromising production aperture and scene composition. Images acquired in 1.77:1 HDTV can be downsampled to 525-line 1.33:1 while maintaining high quality. However, framing will clearly suffer some compromises (see Figure 2). The difficulty is serious, but since productions are now routinely released on standard-definition networks in the letterbox format, the issue is one that consumers will likely understand better than they once did. In Europe, a common compromise is to shoot in 14:9 (1.55:1), which occupies more of the screen when shown on a 1.33:1 display.
At the time of production, aspect-ratio concerns may have a larger effect on live-entertainment programming and sports, which typically have a large number of graphic elements. Tastefully designing graphic treatment so that such a production is not cropped on a 4:3 set, while showing the full intended content on a 1.77:1 set, may require a lot of creativity. For instance, a lower-third name super on a sports broadcast might have a left-justified name on a 1.77:1 screen, but would be badly cropped on a 1.33:1. One solution is to insert different graphics in each broadcast, which might be done in a system in which one control panel switches a show with two video processing frames (and different graphics generators). But this is quite an expensive approach and may only be applicable under unusual circumstances and where budgets permit.
Effectively shooting for HDTV's high resolution and wide aspect ratio also requires different framing. In the first Monday Night Football broadcast that was done simultaneously in HDTV and standard definition, the director shot the program differently for the higher resolution and framing of HDTV. At half time, the ceremony honoring John Elway was shot quite tight (and appropriately) in the 525 broadcast, but much wider in the HDTV program. Both showed the emotion and detail in Elway's face. Each was appropriate to the content and audience. Neither would have been right if it had been up- or downconverted to the reciprocal medium.
Fortunately for our industry, high-quality audio has always been possible during production. But the state of the art has limited our ability to distribute that quality all the way to the home. DTV allows us to carry the (apparent) quality of the production environment to the home — apparent only because it is compressed significantly. But, nonetheless, DTV is of much higher quality than broadcast media has been able to do by other means.
Surround sound is nothing new in our industry. However, most surround sound has been four-channel surround, and DTV allows (but does not require) 5.1 channel surround sound. The channel assignments for Dolby 5.1 (AC-3 coding) are left, right, center, left surround, right surround and low-frequency effects (LFE, which is the “.1” channel). But, not all productions lend themselves to 5.1 sound. For instance, a newscast would derive little additional production value from surround sound, but a football broadcast can be enhanced quite a bit if the crowd audio is placed in the proper spatial relationship in the sound field. The viewer should hear the calls of the quarterback coming from the field, while the crowd sound seems to come from around the listener, and perhaps behind as well. Putting the crowd sound equally in all locations will not enhance the sense of realism that surround sound can project. So, just as a program needs to have a person making visual design decisions (graphics, staging, etc.), a program with surround sound also needs to have someone design the aural portion of the production.
Bringing big-screen values to the small screen involves adjusting to unique demands in the course of acquiring, editing and delivering HD content.
John Luff is senior vice president of business development at AZCAR. To reach him visit www.azcar.com.
Do you have a comment about this article? To tell us your thoughts, click here.