Archives

February 2001 Issue
Feature: Digital-into-Digital Ad Insertion
Integrating Local and Regional Advertising into a Cable TV Multiplex Without Artifacts
By

Recent innovations in digital technology now make it possible for cable operators to run local and regional ads in a digitally compressed multiplex that look every bit as good as the national ads.

A recent Nielsen Media Research study conducted for the Cabletelevision Advertising Bureau revealed that cable viewers pay as much attention to commercials as broadcast television audiences do. That’s good news for cable operators who have the home-field advantage of being able to deliver advertising that targets local and regional consumers. Such retail advertising can generate hundreds of thousands of dollars a year in additional revenue.

However, until recently, the technology for integrating retail advertising into a multi-channel digital video tier has not been possible without introducing loss of quality or artifacts that mar the presentation. Many local and regional advertisers balk at the prospect of running paid commercials that appear shoddy or second-rate alongside glamorous national ads.

Recent innovations in digital technology now make it possible for cable operators to run local and regional ads in a digitally compressed multiplex and have them look every bit as good as the national ads. Terayon Communication Systems, SeaChange International and others recently introduced all-digital ad insertion systems specifically designed to allow cable operators to splice and insert advertising spots into a digital stream without introducing visible image and audio artifacts.

Cable companies across the country are preparing for the day when digital television will completely replace current analog delivery. The displacement will occur for several reasons, not the least of which is the Federal Communications Commission’s mandate that the broadcast industry must eventually convert all analog-based television systems to digital-based systems. The federal agency could call for the return of the analog spectrum as early as Dec. 31, 2006.

Going digital

Digital compression of an analog TV signal dramatically increases reliability of the transmission, decreases distribution costs and increases security. Digital compression technology also increases bandwidth availability, which operators may leverage to accommodate video-on-demand (VOD) or data transmission in such revenue-enhancing applications as high-speed Internet access and interactive (or enhanced) television.

Cable operators typically draw from a variety of local and remote feeds—satellite, video server and local broadcasts—to create a digital multiplex, which can carry more than 200 video channels and additional audio channels. Such tailoring, or grooming the content, of a multiplex to suit the interests of particular subscribers characterized by different social and economic levels, spoken languages, special interests and other local and regional demographics is an effective tool to attract and retain new customers and expand the subscriber base.

Customizing program packages to meet the demands of specific demographics also provides a natural foundation into which cable operators may sell local and regional advertising. In the analog world, operators use simple time-based video switching equipment to insert local and regional ads and programs. Prior to insertion, the video content may be stored in either analog format on traditional videotape or in digital format on standard computer hard drives using advanced digital compression techniques, such as Moving Picture Experts Group (MPEG) compression, then converted back to analog for switching.

With the advent of digital television delivery, operators expect to follow the same revenue model and continue to generate local and regional ad insertion dollars. Unfortunately, the ability to switch digital ads into digital network content has many more complications than a similar solution in analog.

MPEG guidelines used in digital compression techniques may compress the bits in a video stream by as much as 50 to 1. An industry standard for digital video disk (DVD) and some satellite television services, an MPEG compression scheme is nonetheless a lossy method that may slightly degrade picture quality or resolution from that which the digital video camera captures and what appears on TV. However, the quality is still far superior to the average image based on National Television Standards Committee (NTSC) analog video.

At the core of the MPEG compression standard, video streams are compressed using multiple types of frames that may or may not include all of the data needed at any given time. MPEG guidelines use that information that is visually redundant from frame to frame, so it need not be recompressed and may even be discarded.

The standard employs three different frame types:

• An intra-frame (I-frame) uses data from a single original frame and it does not reference any other frames. The I-frame can be decoded independent of other frames in the program and is the only one that can independently deliver a full screen of video.
• A predicted frame (P-frame) compares the data from the original frame and one or more preceding frames. It delivers only the unique frame information identifying a difference between the original and the preceding frame.
• A bidirectionally predicted frame (B-frame) takes this approach one step further and compares the data from those frames that precede and follow. Like a P-frame, a B-frame delivers only partial screen updates. By delivering and updating only the data that has changed since the last frame, the MPEG stream uses much less bandwidth than an uncompressed digital video stream.

Figure 1: Seamless Digital-Into-Digital Ad Splicing System

Challenges

A number of implementation issues arise with digital-into-digital ad insertion, which uses the MPEG standard and will result almost certainly in visible audio or video artifacts.

The insertion point is important because MPEG files contain many codeword sequences embedded in the headers, which are critically meaningful to a decoder and are required in their entirety.

Second, some MPEG-encoded video frames use reference information from subsequent frames, which must be stored and decoded prior to display. If the technician tries to splice between these frames, incorrect reference information will be used to decode the newly inserted MPEG segment. It also is difficult to determine the intended presentation time of a particular frame unit without accounting for the time required to propagate through the rate buffer at the receiver.

A third challenge occurs during data compression. Here the encoder ensures that the rate buffer at a receiver or decoder remains within acceptable limits to avoid buffer overflow and underflow (conditions that arise when a buffer is too full to accept additional data when it arrives, or is empty when additional data is needed). The design of the MPEG compression standard ensures all receivers receiving the same compressed bit stream exhibit identical buffer occupancy levels.

However, files that will be spliced are generally encoded independently. The encoder used for the current file assumes a particular buffer occupancy level at the precise time of the splice, while the encoder utilized for the next program may well assume another. A technician therefore cannot assume that a transition between the two programs will not trigger an overflow or underflow. Unless the two occupancy levels are identical, overflow or underflow becomes possible.

Several solutions are available. One possibility is to decode the incoming digital network feed, and use a half digital, half analog (or hybrid) system to insert an analog ad into a decoded network. Such an approach also requires the purchase of an encoder, a costly piece of equipment used to digitize and compressed both the analog feed and the ad that has been inserted.

From a quality perspective, the phenomenon of generational loss—a discernible degradation in picture quality that occurs each time video content is converted from digital to analog and back—accompanies any approach that relies on currently existing analog-based switching equipment to insert advertising spots into a digital video program stream.

In such a case, the picture quality would degrade when the content changes format from digital to analog and back.

Clearly, only a seamless splicing system may overcome these challenges. Seamless splicing produces an invisible splice between the program material being delivered and a particular ad being spliced into it. Such a process requires encoded ads fit perfectly between splice in-and-out points in the program material and video buffer verifier (VBV) levels match precisely at the transitions.

Seamless splicing systems

Cisco Systems, Harmonic, nCUBE, SeaChange and Terayon Communication Systems are among the vendors that have come up with digital-into-digital ad insertion systems that alllow operators to splice and insert digitally compressed advertising spots into a digital stream without introducing visible image and audio artifacts.

Terayon developed a multi-program splicing system that allows cable operators to splice digitally compressed MPEG files seamlessly into a digitally compressed multiplex containing national ads and custom programming. Terayon’s CherryPicker Ad Splicer allows an operator to insert digital advertising content into a digital feed without decoding the content and converting it to analog format. This system also saves the expense of purchasing an encoder and a decoder for each program to be inserted while preserving original picture quality.

The ad splicer employs a cueing message called DVS-253, which specifies a technique for communicating the precise point at which the ad splicer may splice a compressed bit stream/MPEG-2 transport stream to insert an advertisement. The cueing message must be inserted by a content provider into a program stream prior to uplink. Afterwards, the ad splicer extracts the message and passes it to the ad insertion equipment, which references billing and scheduling information to determine whether to insert the ad. The ad insertion equipment then queues up the ad and the CherryPicker splices the ad, which is embedded in the cue tone message.

Cisco’s 6920 RateMUX statistical multiplexing system can accept digital program feeds of previously encoded advertising from an advertisement video server. A control link between the advertisement insertion scheduler and the Cisco 6920 RateMUX allows for timely splicing of digital advertisements into the appropriate program stream. The real time bit-rate reduction of the digital advertisement to match the bit rate of the source program allows operators to maximize server capacity because only a single copy of an advertisement is stored on the server.

Harmonic’s TransRater TR400 allows operators to splice compressed streams from various broadcast sources. It’s seamless frame-accurate splicing enables providers to insert digital advertising and local programming in real time and without visible picture quality deterioration. Because the TR400 also allows for customizing compressed programming, operators may choose from a number of digital broadcasting sources to select video services and make trade-off decisions between picture quality requirements and bit-rate allocations.

The TransRater system also performs seamless frame-accurate splices and logo insertions. It can splice to an arbitrary frame. At the end of the splice, the system can revert back to the original video stream just as seamlessly.

David Brenner is product marketing manager, DVS, for Terayon. You may contact him at .

 Back to February 2001 Issue

Access Intelligence's CABLE GROUP
Communications Technology | CableFAX Daily | CableFAX's CableWORLD | CT's Pipeline
CableFAX Magazine | CableFAX databriefs | Broadband Leaders Retreat | CableFAX Leaders Retreat