MPEG-H ‘Superior’ for ATSC 3.0 Audio vs. AC-4, Fraunhofer Point Man Says

Despite being denied the ATSC 3.0 prize in the U.S., Fraunhofer still believes in the technical superiority of MPEG-H over AC-4, Bleidt said. “I mean, we’ve demonstrated that it works,” he said of MPEG-H. “But you know, Fraunhofer is more of a technical company. We’re looking to compete on technical superiority, and on sound quality and the technical preciseness of our features, and that doesn’t always win the race.” Dolby representatives didn’t comment.

Though “of course, no one’s ever completely happy” with compromises, “I’ll point out that with MPEG-H in Korea, we’re going to be on the air first,” Bleidt said. “Our expectation is, we’ll be the first system on the air with MPEG-H because we’ll be on the air next year in Korea,” he said of the first ATSC 3.0 global deployment. “And my understanding is that the timetable for the U.S. is at least a year later,” he said.

MPEG-H's Korean deployment prospects for ATSC 3.0 are less riddled with uncertainty than AC-4 for the U.S. because Korea’s ATSC 3.0 broadcast system “will be done on entirely new frequencies,” Bleidt said. “And so, there will not be the transitional issues that we’ll have here in the U.S., which are related to the spectrum auction and the repacking of frequencies and the reassignment of stations, and such. The broadcast architecture in Korea is much more straightforward. The Korean authorities have allocated a completely new set of frequencies just for this new service.”

Bleidt thinks ATSC “did a very thorough job of testing the codecs they were considering” for ATSC 3.0 audio, he said of AC-4 and MPEG-H. “For example, there were listening tests,” he said. There also were “detailed evaluations of the features of each codec in very precise demonstrations that were highly intrusive,” he said. “We at Fraunhofer, we went to Dolby, we sat in Dolby’s listening rooms and laboratories, and watched all their features being demonstrated.”

In turn, Dolby representatives “came to our labs, listening rooms, our theater, in Erlangen, and they watched all of our features being demonstrated,” Bleidt said of the visit to Fraunhofer’s German headquarters. “There was a very thorough and intrusive process in terms of meeting all of the specific features and requirements that the ATSC wanted to do.” That included a special ATSC 3.0 event last summer in Atlanta, where Dolby and MPEG-H felt “compelled to make the most elaborate demonstration possible,” Bleidt said. “We put a complete TV network on the air, all the way from a remote truck to the living room, and so did Dolby.”

At the end of “all of that evaluation, objective testing, feature checking, and a great deal of discussion about what’s needed in an audio system, we came up with two systems, MPEG-H and AC-4,” Bleidt said. “The MPEG-H system was technically superior in some ways,” he said. “Our audio quality is a little better. We can do some features in a little nicer way. We have some new technology in our system that’s not in AC-4.”

But AC-4 “is coming from Dolby, which is a well-known supplier here in the American market and is one that the broadcasters are familiar with,” Bleidt said. “It then became time for the ATSC to make a decision, and as I’ve told people before, some broadcasters wanted to stay with what they knew and were very adamant about having Dolby be there.” On the other hand, “some broadcasters were excited about the potential for MPEG-H” because of its technical superiority, plus also the fact that it’s an “open standard, with more choice of suppliers, and that resonated with them,” he said.

Then there were “the consumer electronics manufacturers" within ATSC, "who unfortunately have to pay for all of this new technology,” Bleidt said. “They were interested, obviously, in minimizing the cost of producing new TVs,” he said. “And so they wanted only one audio system to be selected, because they didn’t want to pay for more than one. After a great deal of discussion, that’s how this compromise was reached.”

Though having been denied selection as the U.S. system, there “most certainly” will be world regions of opportunity where MPEG-H could land adoption as the audio codec for next-gen TV broadcasting, Bleidt said. Both the DVB broadcasting system in Europe and ISDB in Japan and South America “will undoubtedly contemplate new audio standards in the next few years -- not next month, but in the next few years,” he said. “So those are opportunities for the audio companies like ourselves to introduce new technology, and I would expect MPEG-H would be a serious contender for those standards adopted by other countries.”

Achieving economies of scale for MPEG-H-based TVs in Korea “is really not a factor in any of this,” Bleidt said. “For audio, I don’t consider this to be a real issue in reasonable production volumes,” he said. “People have to understand that audio codecs these days -- both ours and our competitors’ -- they’re all software. It’s not like you have to build a special chip to receive MPEG-H. The same class of hardware that you would use to decode one system will basically decode the other system, and this is all software that runs on embedded processors in a chip. So when you speak of economies of scale, it’s really just a question of amortizing development and integration costs to take the software from Dolby or Fraunhofer and integrate it into the rest of the software architecture of the TV. Of course, that takes effort, but it’s not like it takes 1,000 man-years to do that. It’s a reasonable engineering effort.”