WASHINGTON — Despite pressure from federal regulators to resolve problems in connecting cable set-top boxes to digital TV receivers, industry officials concede that first-generation sets will not be equipped to receive DTV broadcasts via cable. The lack of digital interfaces will force early adopters this Christmas to install the '90s version of rabbit ears on their sets to capture digital signals.

"First-generation [digital TV] is a problem," said Bill Wall, chief scientist at set-top maker Scientific-Atlanta. In a nutshell, the first digital TVs and set-tops will lack the 1394 interfaces needed to receive digital broadcasts, which are slated to begin in November.

FCC chairman William Kennard had called on cable operators and consumer-electronics manufacturers to produce IEEE 1394-enabled digital TV sets by November 1999, with the expectation that a firm 1394 interface spec could be achieved a year before that deadline.

"I see no reason why this should not be complete by Nov. 1, 1998," Kennard said in his Aug. 13 letter to the National Cable Television Association and the Consumer Electronics Manufacturers Association (CEMA).

Thus far, however, the cable industry has not agreed to deliver terrestrial digital-TV broadcasts in an unadulterated form and incorporate high-definition decoders in cable set-tops. That's because high-definition transmissions will eat into bandwidth allotted for conventional broadcasting.

CEMA officials did not respond to requests for comment. Key players in the consumer-electronics industry nevertheless claimed last week that they are close to settling IEEE 1394 digital-interface issues within Kennard's November 1999 deadline. A compromise on the interface issue among manufacturers, cable operators, Hollywood production studios and other key players is expected this fall, industry and FCC officials said. If all goes well on the 1394 front, systems that could handle DTV signals from cable operators could be ready for consumers by fall 1999.

In place of the desired 1394 interface, the DTV receiver sets available this fall will use an analog component video-the same kind of a connector used in professional video equipment-to connect the DTV receiver and the DTV decoder box.

Other interim solutions are in the works. For instance, cable operators could carry the 8-VSB digital TV signal directly on their systems. Another alternative would be for set manufacturers to include quadrature amplitude modulation (QAM) demodulators in DTV receivers.

Scientific-Atlanta is pursuing its own alternative solutions, Wall said. One is a high-definition version of its Pegasus set-top that would use an analog component interface to connect with a DTV receiver. The company also expects to unveil a decompression scheme for its digital set-top early next year.

Yet 1394 may be the least of the problems faced by designers of digital TV systems. Other unresolved issues-what some engineers are calling the "dirty secrets of DTV"-center on adequate definition of conditional access systems for terrestrial digital broadcasts, including strategies for implementing such systems as removable conditional-access modules in DTVs or set-tops and mechanisms for ensuring copy protection at the set-top/module interface.

Conditional access involves "complex implementation issues," said Chris Adams, vice president of marketing for the Consumer Network Products Division of C-Cube Microsystems.

With copy protection and encryption issues unresolved, the first-generation DTV receivers scheduled to debut this Christmas appear to be designed on the assumption that viewers are more likely to receive over-the-air DTV broadcasts via antenna than to receive signals via cable. Further, because they will lack terrestrial-DTV conditional-access systems, first-generation DTV tuner/decoder boxes, priced as high as $1,700, will be doomed to obsolescence within a few years, when network TV studios start broadcasting some programs using encrypted DTV signals.

Complicating matters, there has been little progress on the larger issue of how cable operators will relay digital-TV broadcasts to their subscribers. The FCC launched a review of the digital-TV "must carry" issue in July. The first round of industry comment is due Oct. 16. The FCC's notice seeks comment on issues related to the transition to digital TV, including digital equipment compatibility. At this point, it is not known whether cable operators will "pass through" digital broadcast signals in their original broadcast format or whether they will down-convert them to certain formats at their head ends. It is not even known yet whether the FCC will require cable operators to carry a given number of terrestrial DTV channels in specified resolutions under the pending"must-carry" provisions.

Observers said Kennard's letter to CEMA and cable operators appears to recognize that nailing down the 1394 digital-interface issue is the first step toward ensuring compatibility between cable set-tops and DTV receivers.

In including a 1394 interface on both the cable set-top and DTV receiver, it is assumed that the set-top will be responsible for receiving and demodulating QAM-modulated DTV cable signals and decrypting premium-service programs. The set-top will not, however, be able to decode high-definition TV signals. Current digital cable set-tops, equipped with only MPEG-2 Main Profile @ Main Level decoders, will need to pass on the compressed MPEG stream to DTV receivers over the 1394 interface so that the receivers-presumably equipped with MPEG-2 Main Profile @ High Level decoder chips-will be able to decode HDTV signals.

In his letter, Kennard specifically asked the industries to achieve consensus on "additional higher-level protocols and standards that must be defined for 1394-enabled devices to perform functions such as copy protection and the on-screen display of programming information."

Copy protection and on-screen display over 1394 are separate issues and have been explored by two different industry working groups, according to Scott Smyers, vice president of the Interconnect Architecture Laboratory of Sony Electronics U.S. Research Lab (San Jose, Calif.).

Smyers, who chairs the digital-transmission discussion group within the cross-industry Copy Protection Technical Working Group, played a key role in forging industry consensus on development of a secure path for digital transmission of copyrighted material among PCs and other digital consumer devices via a 1394 transport-encryption algorithm. The focus of that effort is to protect a bit stream from illegal copying once it is descrambled within a set-top and exposed to external devices (PCs or digital VCRs) through a high-performance serial bus such as 1394.

After nearly a year of discussions, five companies-Sony, Hitachi, Matsushita, Intel and Toshiba-earlier this year agreed to merge their separate proposals for the 1394 transport-encryption algorithm. The final specification, completed a few months ago, accommodates a number of last-minute requests by the Motion Picture Association of America (MPAA, Washington).

The five companies now are setting up an independent licensing entity, according to Smyers. Intel has volunteered to act as an interim licensing authority and is accepting requests from chip vendors and system companies wishing to implement the transport-encryption algorithm for 1394.

Smyers called 1394-related copy protection "a done deal," adding, "I have not met anyone at the MPAA who does not endorse this encryption scheme over the 1394 interface."

But others stressed that delivery over the 1394 interface of uncompressed graphics-overlay information along with compressed MPEG video streams is a different "deal" that's not yet done. The 1394 Trade Association's A/V Working Group is working toward a solution.

The graphics issue is important for service providers and set-top vendors that seek to create their own distinctive look and feel for on-screen displays of electronic programming guides and other new digital services. Without a commonly applied method for sending raw bit-map graphics information from set-tops to receivers, DTV-set makers may be forced to turn out multiple digital-receiver types to accommodate various set-tops, warned A/V Working Group chairman Smyers.

The group is considering ways to send graphics commands over the 1394 interface and to format data. Also needed are multiple definitions of what pixels should look like when graphics-overlay data is sent over 1394. "We may need several options there, rather than one. While some may wish to implement just 4-bit, 16 colors, others may want to use a much larger number of bits and colors for their [programming guides]," said Smyers.

The A/V Working Group is not expected to complete its work until January. Responding to Kennard's letter, the 1394 Working Group struck a deal with the Electronic Industries Association subgroup defining the digital-TV interface to put the effort on the fast track, Smyers said. The EIA agreed to include the A/V Working Group's plan in its spec in a bid to meet the FCC's deadline.The A/V group will iron out the technical wrinkles for the final specs.

Questions nonetheless remain about how soon manufacturers can move once the interface issues are settled, noted Scientific-Atlanta's Wall.

For example, it is not known how many chip and systems companies have licensed the 1394 encryption algorithm. Smyers said including the encryption algorithm on a 1394 chip requires "adding 10k to 12k gates." The task may take silicon vendors several months to complete, but it's not monumental. "It won't take a year," Smyers said.