Broadcast and Production Embrace 3-D
Content providers and consumer-electronics companies are already behind the move to 3-D. Now, professional broadcast and production vendors are joining in to help provide critical mass.
by Chris Chinnock and Matthew Brennesholtz
IN APRIL, the National Association of Broadcasters (NAB) held its annual exhibition in Las Vegas. The NAB show is geared to the latest technology in mobile, terrestrial, cable, satellite, and Internet broadcasting, and encompasses the creation of content, post-production, live-broadcast events, and the management and delivery of content to consumers' homes. Stereoscopic 3-D was certainly the most visible technology innovation at this year's show. As a result, it now seems clear that the current wave of interest in 3-D for consumers will continue. All hardware and software suppliers to the broadcast and production communities are now on-board and offering products to serve the 3-D segment.
The change in attitude toward 3-D from last year's event to this year's was profound. In 2009, a handful of companies showed 3-D products and many noted that they were studying 3-D. But a year later, every aisle at NAB had 3-D products and services and nearly every company was talking 3-D (see Fig. 1). Some of this seemed more talk than fact, but the fever has taken hold. As one of Insight Media's analysts, Bernard Mendiburu, noted, "It seemed like 5 NAB years had passed in a single year." That's how rapid and dramatic the shift has been.
Fig. 1: The 3D Hero camera from GoPro belongs to a line of professional-level, wearable, HDTV and action-sports cameras. It is expected to be available later this year.
Factors Behind the Rise of 3-D in Broadcasting and Production
There seem to be multiple factors for the sudden change in emphasis. One is that 3-D content is being created and pushed by the major Hollywood studios, and this effort is meeting with great success (and profits) in theaters. Studios are also pushing to move this same 3-D content into the home because a large percentage of Hollywood revenue comes from non-theatrical sources. More and more 3-D content will also come from ESPN, Discovery, and a host of new channels and content producers. The 3-D pipeline is filling, so this piece of the infrastructure is maturing rapidly.
Next, TV makers have introduced 3-D TVs with many advanced features – Internet connectivity, LED backlights, Widgets, Skype video conferencing, and more – not just 3-D capability. Consumers will buy these TVs for 3-D and the other features, creating a growing installed base of 3-D-capable TVs. The Blu-ray 3-D specs have been established, as have specs for delivering 3-D over HDMI cables. So the pieces are in place for 3-D in the home. Pricing is at the high end of the range for these products, but the premium for a 3-D-capable TV compared to the premium for an HDTV over a standard-definition TV 10 years ago is significantly lower.
Finally, it is vitally important to be able to deliver 3-D content via cable and satellite services. Having made a costly investment in upgrading the infrastructure to support digital and high definition, this part of the chain does not want to make a similar investment for 3-D. For cable and satellite service providers, squeezing 3-D into the existing infrastructure is a must if early 3-D rollouts are to be enabled. Fortunately, this is possible and exactly what is being done. The two high-resolution stereo images are filtered (sampled or decimated) and then packed as side-by-side or top/bottom images in a standard video frame. This is the so-called frame-compatible approach. This signal can be processed, compressed, transmitted, and decoded like any other 2-D video signal. For the most part, some minor firmware upgrades are all that are needed to enable transport and delivery of 3-D signals to a consumer's 3-D TV. Only when the content gets to the TV does the TV recognize it as 3-D and process it to display a 3-D image. For an overview of the flow of 3-D content into homes or movie theaters, see Fig. 2.
Fig. 2: 3-D content flows from acquisition though post-production and broadcast formatting to delivery in homes and theaters. Image courtesy Harris.
There are two main disadvantages to this approach, however. First is the obvious fact that this decimation of the image significantly reduces image quality. With some types of proprietary encoding, such as that provided by Sensio or RealD, the image-quality reduction is not really significant, but then you need the proprietary decoder at the receiver. If you do not have the right decoder in your 3-D TV, the signal will still be viewable in 3-D, but there will be a significant loss in image quality. And, as we have seen from some early demonstrations such as the Master's Tournament in HD 3-D, the bit rate devoted to the 3-D signal is important, with 18–20 Mbps clearly being the best choice for HD image quality.
The second problem is that the frame-compatible approach doubles the required bandwidth – but in a not-so-obvious way. Since this frame-compatible 3-D image is not viewable as a 2-D image on a 2-D set (a side-by-side or top/bottom image is seen), the cable company must have a second channel to deliver the 2-D signal. For the near future, broadcasters are planning only a few 3-D channels to go along with their many 2-D HD channels so we do not see separate 3-D channels as a major issue for now. If 3-D broadcasting becomes common, it may be more of an issue, but alternative solutions are being developed now.
The bottom line is that all of the pieces of the value chain to create and sustain a 3-D TV service for consumers are in place. But is there a business model to support this? It would appear that there is.
For example, TV is supported ultimately by subscription fees and advertising. While consumers might pay more for a 3-D video on demand of a pay-per-view event, as they do in the theaters for a 3-D movie, advertisers must also embrace 3-D. Panasonic and others are helping to prime the advertising pump by injecting some big bucks to sponsor the establishment of 3-D services from DirecTV. This allows DirecTV to help pay the bills while it figures out the service and gets advertisers onboard, and while consumers buy 3-D TV sets. This is a brilliant risk-reduction strategy that jump-starts the market.
Companies in the content creation and broadcast arenas can also manage risk. One theme we heard at NAB from the equipment vendors is that the broadcast industry is not buying 3-D equipment. Companies are buying digital HD production, post-production, and broadcast equipment – but requiring that it be 3-D ready or field-upgradable to 3-D with software or firmware changes. In some cases, vendors describe their 3-D path as adding a second identical piece of equipment to handle the two signal paths. But vendors with only 2-D HD equipment not suited for upgrade to 3-D seem to have been getting the cold shoulder. This shows that the infrastructure upgrade to support 3-D is dovetailing well with digital HD production and broadcast needs, lowering the risk of support. And adding 3-D capability to most equipment is not such a big deal, as most vendors have been able to do this in 1 year. Future proofing, editing, and broadcast control suites for 3-D is nowhere near as intensive an investment as was the HD upgrade.
3-D Production and Broadcast Equipment
There are areas that do require dedicated 3-D equipment, however, including cameras, camera processors, and displays. For example, stereoscopic 3-D cameras typically use two sensors and two lenses to capture the left/right image pair. These cameras can be configured in a side-by-side manner or in a beamsplitter arrangement with a precision optical element positioned between the two cameras, which are orthogonal to each other. These 3-D rigs need to perform all the 2-D camera functions, plus some 3-D ones. The distance between the cameras usually needs to be varied to obtain the right 3-D perspective. The cameras may have to be "toed in" just as our eyeballs toe-in to look at closer objects. In addition, the camera motion, stability, and matching of lenses and other parameters are all very important for good 3-D acquisition. Last year, there were perhaps a half-dozen 3-D camera-rig models available worldwide. At NAB this year, there must have been close to two dozen and at a wide variety of price points. This is a needed part of a healthy ecosystem.
For 3-D displays, two types are being used by the professional community: one is the beamsplitter type, in which two flat-panel displays are positioned at right angles to an image combiner. These are often boxy monitors but are popular for monitoring 3-D images during movie or event shooting. Passive polarized glasses allow both images to be seen simultaneously, but by different eyes.
The second type is the X-pol or Micro-pol. This display features a sheet of retarders that is laminated over a flat-panel display. The sheet converts the polarization of the panel such that even rows have one circular polarization state and odd rows have the opposite state. Passive polarized glasses then allow the viewing of the 3-D image. These specialized 3-D monitors are not likely to be used for 2-D work, so they are an investment upgrade.
Note that both of these approaches use passive polarized glasses (see Fig. 3) and not the active shutter glasses favored for home-based TVs. For the home environment, TV manufacturers did not want to put the 3-D cost into the display, but into the glasses that the consumer buys, thereby making the TV pricing more competitive. In the professional world, long hours of use and fears about interference in the image quality by the shutter glasses mean passive approaches are preferred. In addition, in the professional production, post-production, or broadcast environment, there are likely to be multiple 3-D displays visible at a time. Passive glasses bypass the issue of synchronizing all these displays so the operator sees the correct image on all displays at all times. These users will pay more for the display and less for the glasses because this technology better suits their needs.
Fig. 3: This viewfinder from Miracube for a 3-D camera system is based on micropolarizer technology and passive glasses.
It should be noted that anaglyph glasses are also used in editing and broadcast suites and will continue to be in use for some time, we suspect. There are two reasons for this. First, any 2-D display can be used to view 3-D with anaglyph glasses. Second, it provides an easy and effective way to monitor the 3-D image even without the glasses on. When a red/blue anaglyph 3-D image is viewed without glasses, it provides a quick and accurate indication of parallax. If the blue fringe is to the left and the red fringe is on the right, parallax is positive and the object will appear safely behind the screen. If the colors are reversed, parallax is negative and out-of-screen effects are occurring, potentially causing viewer eyestrain.
We do not see 3-D monitors being used for critical activities such as color correction or even for editing. For the most part, editing of 3-D content will be done in 2-D and checked in 3-D. Color grading will also be done using Grade-1 2-D monitors for the foreseeable future. However, editing and color-correction software and hardware are coming to market to make it easy to replicate the edits and color correction from one image to other images, thereby reducing the production/post-production time for 3-D content.
"Workflow" is a key buzz word in the post-production and broadcast industries. It refers to how content flows from one stage of the process to the next – ideally easily and seamlessly. While 3-D workflow is not yet as seamless as 2-D workflow, it is getting there.
New processors and encoders may be needed to support 3-D. The processors can manage things such as camera metadata, which include zoom, focus, position, toe-in, convergence, and color space. Some of these parameters are all needed by 2-D workflows, but new parameters are needed by the 3-D workflow to help assure a well-aligned and well-composed 3-D shot. Encoders will need upgrading to support the two high-quality 3-D images and to filter and pack them for transport in various parts of the distribution chain. There may also be proprietary encoders involved to improve image quality.
Digital Production and Digital Products Are Aligning
In conclusion, NAB 2010 was not the NAB show at which 3-D became common. Rather, it was the NAB show at which everyone recognized the need to be ready for the arrival of 3-D broadcasting in the near future. One of the promises of digital systems is their easy upgradability for new applications. The dramatic 1-year growth of available 3-D equipment and 2-D equipment future-proofed for 3-D is a sign of the industry making good on this promise.
Comparing NAB 2010 to CES 2010 is an interesting exercise. CES provides insight into what consumer-electronics companies are thinking, especially with regard to 3-D TVs, which were a major theme of CES 2010. NAB reflects the views of broadcast, production, and post-production. These views from both areas of 3-D are now aligning. Insight Media and others are forecasting strong sales of 3-D TVs, and it appears the content and infrastructure will be there to support those sales. It now seems clear from NAB that more 3-D content will be created and the production and distribution mechanisms will be in place to support a ramp up of 3-D. However, the last and most important issue is consumer acceptance of 3-D. We think 3-D will be accepted by consumers for event-driven viewing, but realistically, it is too early to know the answer to how much other such viewing they will accept. The first 3-D TVs and 3-D services are starting to roll out now, so the answer will become clearer in the coming months and years. •