A number of readers have commented via e-mail about the “Beyond 1080p” article in the Nov. 18 “Transition to Digital.” The topic seems to have hit a nerve. Readers raised questions about implementing a 1080p production infrastructure and methods of distribution.
John Luff, television technology consultant and SMPTE Fellow, expressed observations about the viability of production infrastructures being able to produce content in 1080 60p, let alone 2K resolutions. He pointed out that:
….8-bit, not 10-bit sampling, … for production would be inadequate [and] 4:2:0 coding will sustain for home use into the future, which might be a stretch when people commonly get higher resolution displays… progressive scan formats compress better as well, which probably makes up for the 8-bit/10-bit differential plus a little, but only for home distribution, not production. There is a sentiment on the part of many in production that 12 bit would be better, especially when many output formats including 1080p60 are considered.
…the complete format [of] SMPTE 292M is actually 2200 x 1125 lines. The space between active picture (1920*1080*10*2= 1.24Gb/s) and the total bandwidth (2200*1125*10*2= 1.485Gb/s) must have room for all metadata and audio, which it obviously does. 1920 x 1080p24 production format is .995Gb/s for active picture, and making that p60 raises the bar to 2.488Gb/s.
And of course, Mr. Luff is absolutely right. This is precisely the point. The theme of “1080p and Beyond” was to point out that broadcast production infrastructures will have difficulty supporting 1080 60p, while Digital Cinema production has already moved up to 2K and 4K production. So, the movie industry can create content in a format, 1080 60p or 2K, that broadcast production cannot.
Implementation of dual-link 3Gb/s distribution methods for 1080 60p would reduce the capacity of an existing HD infrastructure by 50 percent. SMPTE has issued 3Gb/s standards (424M, 425M) that can be used for intrafacility 1080 60p SDI distribution over a single coax cable. An alternate method could be to work in the compressed domain, with 1080 60p content compressed at 200Mb/s data rates or higher. This format could be distributed using SDTI, ASI or over the media network (TCP/IP). In this way, 1080 60p content will survive the editing and production process, and produce acceptable, artifact-free video.
Broadcasters are concerned with converting their production infrastructures to support 1080i (or 720p) HD content. Any thought of expanding production to 1080 60p or 2K is well out of mind. Mr. Luff’s observations support the article’s premise that the Digital Cinema industry may be the only source for true 1080 60p native format content.
Delivery channels and display resolution
The true benefit of 1080p display is not 60p necessarily, but the fact that all 1920 pixels are present on each line. And if 2K displays really come to exist for the home theater, both formats need a delivery mechanism.
An e-mail from reader Joe Gombos of JBA Consulting Engineers asked for more information about how existing delivery channels could support 1080p.
A potential OTA delivery solution could be to use the ATSC A/53 High Data Rate (HDR) mode, which employs a 38.78Mb/s Transport Stream (TS) and 16-VSB modulation. This would facilitate squeezing a 1080 60p, MPEG-2, 4:2:0 color sample structure into 6MHz. Of course, there is a more stringent signal to noise requirement of 28.3dB, as compared to 14.9dB for 8-VSB with a corresponding decrease in coverage area.
A cable system can go to 256-QAM, while DBS would have to allocate a complete satellite transponder. Both methods would eat up valuable channel capacity. In addition, the resulting signal could be less robust and therefore result in less image stability.
Telco IPTV and FiOS delivery systems are approaching 100Mb/s. And although 40Mb/s may or may not be a practical, guaranteed bit rate to the home now, in the future, it may prove to be sufficient for 1080p given that a High Data Rate TS is at 38.78Mb/s, provided that it is a guaranteed, sustainable QoS.
Broadband cable Internet connections have not yet reached sustained data rates at the 40Mb/s level. So, a download mechanism may do the trick; however, it will take 12 hours at 10Mb/s to download a three-hour 1080 60p movie. This may prove to be unacceptable. Cable operators may have to turn to a switched video implementation. In this way, similar to VoD, only the one stream of content requested by a customer is delivered. HD VoD and PPV capabilities exist, but again, going to 1080 60p will take up channel capacity and limit the amount of content that can be delivered. Maybe the near-video-on-demand model will work, starting 1080p content every 15 minutes.
With the emergence of more efficient compression engines such as AVC and VC-1, however, bit rates for 1080 60p may be reduced to manageable numbers in the near future. Will AVC/VC-1 compress native 1080 60P content sufficiently to enable delivery over existing channels? Will a Super HD-DVD come into existence, or will quad-core media PCs handle native 1080p?
Regardless, content availability will still be the fundamental issue. So, as discussed last month, the digital cinema production industry may be the only source for content in these higher resolution formats.
Technology necessitates change
In the end, naturally, a 1080p display/DTV receiver will be required. This necessitates new 16-VSB/256-QAM demodulators. The High Data Rate TS will have to be parsed. New decompression engines will be needed. And for 2K pixel grids, new display manufacturing techniques or technologies may be necessary. Seeing how true 1920 pixels per line capabilities are just now being introduced, I wouldn’t hold my breath waiting for higher display grid formats and advances in processing circuitry.
By now, just about everyone in broadcasting realizes that “broadcasting” isn’t just about broadcasting any more. Multichannel delivery and multiformat production have completely altered the business. There is no “user manual” that can keep up with the speed at which the industry is changing. The responsibility rests with industry professionals to find the solutions to these engineering problems as they arise. As some have said, this is a new engineering discipline born of the union of broadcast engineering and information technology.
Happy Holidays to all!