Ateme squeezes more from MPEG-2, targeting US ATSC capacity crunch

By Barry Flynn, Contributing Editor

French video compression specialist Ateme has developed a new high-efficiency MPEG-2 codec, HE-MPEG2, for US broadcasters facing reduced capacity after the FCC’s broadcast incentive auctions.

Currently, according to the company’s VP of Business Development, Yaron Raz, a US broadcaster can expect to transmit around one HD DTT channel plus one or two SD ones within a single MPEG-2 ATSC 1.0 multiplex.  “What we are offering with HE-MPEG2 is basically to do two HD [channels] and two SD [channels], and that would be a combination of 1080i or 720p, so either case,” says Raz. “Some operators are looking to get even more than that, to get three 720p HD channels [into a single ATSC mux].”

The FCC’s broadcast incentive auction, which began on March 28 and is set to continue into the summer, comprises two separate elements: a ‘reverse auction’, which will determine the price at which broadcasters who choose to participate will voluntarily relinquish their spectrum usage rights; and a forward auction following the first, which will determine the price companies – probably mostly telcos – are willing to pay for flexible use wireless licenses.

“Obviously, the [broadcasters] that give up spectrum don’t want to go out of business,” observes Raz. The expectation is that they would therefore work out deals with stations that have elected not to participate in the auction, to share their frequencies, he suggests.

This is the market HE-MPEG-2 is targeting. The technology offers the possibility that, by increasing multiplex efficiency, the two parties to such a deal could maintain most – if not all – of the channels they had each supported before the auction.

Raz explains that HE-MPEG2’s ability to achieve this is partly due to switching from dedicated hardware-based MPEG-2 encoding to off-the-shelf CPUs running its proprietary hardware, and partly due to leveraging “some of the ideas that were developed around MPEG-4 and HEVC and re-applying portions of them back to MPEG-2.”

The resulting jump in multiplex efficiency requires no change in ATSC receivers, which ‘see’ the HE-MPEG2 mux as an ordinary MPEG-2 stream, just one that happens to contain more channels.

Migrating to the next-generation ATSC 3.0 standard, which is significantly more efficient than ATSC 1.0 and is now essentially complete, might appear to offer an alternative solution to the capacity crunch. However, points out Raz, “it’s a completely new protocol, it’s based on scalable HEVC, not on MPEG-2, and requires a completely new set of receivers. In order to transition to ATSC 3.0, you need all the devices on the other end to change, so obviously this is not something that broadcasters can do today.”

Raz envisages that HE-MPEG2 will be supplied to broadcasters as an upgrade to their old hardware-based MPEG-2 ATSC encoders, in the shape of an off-the-shelf CPU (which either Ateme or the broadcaster can provide) running Ateme’s HE-MPEG2 software.

This new encoder will be capable not only of broadcasting ATSC 1.0 channels, but also be able to stream OTT versions of them as required. It will also be able to handle ATSC 3.0 – when the time comes – as a simple software upgrade.

While HE-MPEG2 is being initially targeted at the US market, Raz recognizes that considerable quantities of legacy MPEG-2 equipment exist elsewhere which could benefit.

One obvious market is South Korea, an ATSC 1.0 territory which will probably switch to ATSC 3.0 ahead of the USA. Meanwhile, notes Raz, US cable still runs on MPEG-2, even though around 60% of cable set-top boxes contain dual MPEG-4/MPEG-2 chipsets. “At some point it makes sense to make that switch [to MPEG-4], but for a lot of them, it’s not today,” he points out.

“Today most cable operators usually put about three to four HD channels on a QAM [carrier]. With HE-MPEG2 they can get to about five or six HD channels per QAM carrier – so it’s about a 50% increase in the number of channels they can carry.”