ATSC 3.0 Framers Delay HDR System Choice for 2 Months to ‘Get It Done Right’

ATSC’s S34-1 ad hoc group was to have chosen an HDR technology by July 31 when the candidate-standard period on the ATSC 3.0 video spec expires and next gets elevated to the status of proposed standard, ATSC’s annual broadcast conference was told last month (see 1605100047). ATSC 3.0 video was elevated to the status of A/341 candidate standard in December without HDR included (see 1505130058). Six HDR proposals are now being considered for selection as the ATSC 3.0 HDR spec, ATSC has said (see 1605200031).

S34-1 still plans to go ahead with its comparative HDR demonstrations on the various proposals next week at CBS Labs in New York, said Dave Siegler, vice president-technical operations at Cox Media Group, in Q&A during the Thursday webinar. “Certainly, there’s a lot of attention and focus and excitement about HDR” in ATSC 3.0, Siegler said. “The folks that are working on the standard want to make sure that we do it right and do it once,” Siegler said. “So they’re taking a bit more time to work through that and do those evaluations next week and see what those tests yield.”

When ATSC 3.0's framers “initially envisioned” the next-gen standard, “HDR and wide color gamuts and high frame rates weren’t conceived of back then,” Siegler said. “So sometimes you gotta zig, sometimes you gotta zag. It’s taking a bit longer, but we’re going get it done right.”

Since there was wide consensus within ATSC to pick the H.265 codec for ATSC 3.0 video, “we did not go through comparative testing in a rigorous, subjective fashion" on the A/341 candidate-standard document "like we did with audio, where we had competitive systems proposed,” said Skip Pizzi, NAB senior director-new media technologies, also in Q&A. “However, now we’ve got competitive HDR proposals, so we are trying to do a more rigorous evaluation of just those systems and components,” Pizzi said. “And it’s a complex ecosystem that we have to think about, how that would fit within, whether or not metadata can be passed along, and so forth, and what if it doesn’t come through. So there are a lot of permutations to consider.”

That the end of the candidate standard period on A/341 was extended to Sept. 30 “does not mean that there’s necessarily a delay,” Pizzi said. “It just gives the standards group a little more time to work on getting it right.” A/341 can be moved quickly and at “any time” between now and Sept. 30 to the “next step” of a proposed standard, Pizzi said. “We just didn’t want the candidate standard period to expire, because if it does, it goes back to a working draft and then would have to be republished. It just gives a little more of a runway to land the plane on.”

ATSC 3.0's framers still expect the delay “to basically only amount to a matter of weeks in terms of giving us a little more flexibility to get this HDR and some other 't's crossed and 'i's dotted in the video standard to get it right,” Pizzi said. “Once it goes to proposed standard, it has to be balloted, and then any substantive change made after that has to be reballoted, whereas while as a candidate standard, it’s still flexible to make a lot of changes. So that’s where we’re at with that right now.”

That A/341 still showed a July 31 candidate-standard expiration date when we checked the ATSC website late Thursday suggests that the decision to delay to Sept. 30 was very recent. In fact, the decision to delay wasn't so much recent as it is unofficial, ATSC chief Richer emailed us Thursday. The "status" of the A/341 video document "has not changed" in that its candidate-standard period "currently expires" July 31 and so is still posted as such on the ATSC website, Richer told us. "It is possible that Technology Group 3 may decide to extend the period during its meeting on July 14th," Richer said of the TG3 committee that's supervising overall ATSC 3.0 standardization activities. "Unfortunately, the slide Skip used incorrectly indicated the period has already been extended. HDR demonstrations and testing are still on the calendar for next week."

Broadcasters are “looking at 4K as the part they might not actually transmit” under ATSC 3.0, Pizzi said on the webinar. They’d rather “do something in 1080p with HDR and wide color gamut” at 10 bits, he said. A big reason is that a 4K signal is estimated to consume 300 times the bandwidth of 1080i, though HDR consumes only 20 percent more and higher frame rates of 60 frames a second only 25 percent more, he said. “So at 60 or 120, you still do not use a whole lot more bits in the channel,” he said of 60- and 120-Hz frame rates. By comparison, “when you actually send native 4K content, you do,” he said.

“We’ve been doing some experiments where you take original 4K content, down-rez it to 2K 1080p, and then transmit it to a system that has an upconverter,” Pizzi said. “It’s quite difficult to tell the difference” in most cases between that down-rezed picture and the one in native 4K, he said. Taking that approach will be “saving a lot of bandwidth in the process,” he said. “So this will be one of the possibilities that will be allowed in ATSC 3.0, and the broadcaster can choose all of these elements, sort of a la carte.” Pizzi’s prognostication that many broadcasters will shy away from 4K transmissions under ATSC 3.0 differs significantly from Samsung expressions of optimism a year ago that many by and large would use ATSC 3.0 to beam the highest possible resolutions to consumers (see 1506170046).