For many years, ‘interfacing’ a video signal meant plugging in a yellow RCA or silver BNC connector that carried composite video. As picture resolution went up, computers became commonplace at work and home, and the term ‘progressive scan’ entered the lexicon, we saw the birth of S-video and then component video (YPbPr and RGB).
So we adapted, building switching and distribution gear that could handle one-, two-, and three-wire formats. All was well and good…until ‘digital’ made its grand entrance about 15 years ago.
Now, we have digital versions of component and RGB video, starting with the Digital Video Interface (DVI) and moving to High Definition Multimedia Interface (HDMI), DisplayPort, and the new superMHL interface that (according to the MHL Alliance) will start appearing on televisions as soon as December.
If I’m a consumer, I mostly don’t care about any of this. As long as I can plug in my set-top box, Blu-ray player, and other gadgets with the right cables I can find at Best Buy, this is just a bunch of alphabet soup.
However, if I’m an integrator (consumer or commercial), then I care VERY much about where all of this is heading. And if I’m paying any attention at all to the growing market for 4K and UHD, then I’m rightfully concerned about the impending problems with interfacing these signals.
*Even though HDMI 2.0 was announced in September of 2013 – TWO FULL YEARS AGO – virtually no manufacturer in the pro AV space supports this interface on their switchers and distribution amplifiers. Instead, the vast majority are still providing version 1.4 while claiming these products are “4K compatible” or “4K ready” because version 1.4 is just fast enough to pass an Ultra HD (3840×260) signal at 30 Hz with 8-bit RGB color. That’s setting the bar kinda low, isn’t it?
*Some computer manufacturers don’t even support HDMI, like Apple (DisplayPort) and Lenovo (also DisplayPort). So, now you have to carry dongles everywhere you go?
*HDMI 2.0 arrives hand-in-hand with a new version of copy protection (HDCP 2.2) which is much more rigorous than versions 1.3 and 1.4. If a valid HDCP key exchange isn’t made within 20 milliseconds, the connection will shut down. Period.
*HDMI 2.0 isn’t fast enough for what UHD is turning out to be – a real departure from 1080p and Wide UXGA, with a move to 10-bit color to support high dynamic range (HDR), wide color gamuts (WCG), and high frame rates (HFR). DisplayPort 1.2 can barely support these requirements; DP version 1.3 and super MHL are better positioned to handle the job.
*The intellectual property behind HDMI and superMHL is owned by the same company – Lattice Semiconductor – and whereas once there were clear dividing lines between the two interfaces (MHL was designed originally for smartphones and tablets), they are now competing against each other. I’ve even sat in on presentations where it was explained that both could exist on consumer TVs. (And why would that make sense, again, when neither interface has been widely deployed to date, and one is clearly an improvement over the other?)
You can imagine what this trend is doing to product designers and manufacturers. Sure, HDMI is a “safe bet” for now, but what if our UHD needs quickly outstrip its maximum clock speed? DP is certainly faster and there appears to be more support for it from computer manufacturers. But super MHL is faster still. Shouldn’t your interfaces at least have a head start on display manufacturers?
This reliance on HDMI has led several manufacturers into a potential trap, investing heavily on signal distribution architectures that may quickly run into a “future-proofing” problem. In contrast; outside the commercial AV industry, everyone from cable TV system operators to broadcasters and telecom operators are busy migrating to an IP-based architecture.
Not only does IP-based architecture have the advantage of being a relatively open system, it also solves many of the speed issues as 1-gigabit and 10-gigabit networks are becoming more commonplace. (Heck, Comcast just upgraded my home Internet speeds to 75 Mb/s on downloads, which is more than fast enough for me to stream 4K content from Netflix and Amazon!)
So, why don’t we do the same in the commercial AV industry? It’s not for a lack of products – there are several companies offering AV-over-IP transmitters and receivers, along with encoders and decoders. I’ve also seen impressive demos of “middleware” used to locate, switch, and play out media assets over IP networks. All of these guys were at InfoComm 2015.
The big players in HDMI-based switching and distribution argue against AV-over-IP for in-room and short-run signal distribution, citing latency and compression issues. Well, we now have a new codec (HEVC H.265) to handle that end of things, and it’s possible to stream video and high resolutions with low latency. (How does 1920x1080p/60 at 1 to 2 Mb/s sound to you? Thought so.)
High latency is often the result of over-compression and heavy forward error correction (FEC). But if video and audio assets are streaming on bandwidth-managed, private IP networks, there isn’t a lot of forward error correction required. Group of Pictures (GOP) sizes can also increase to reduce latency. So latency is sort of a “straw man” argument. (And HDMI 2.0 will have plenty of issues with HDCP 2.2, trust me. Talk about latency…)
As for copy protection; video and audio assets streaming over IP connections have their own security protocols. Practically speaking, what could be more secure than video content streaming directly into a UHDTV, through an Ethernet connection? And you don’t even have to plug in a cable to make it work, unless you use a wired Ethernet hookup. Bandwidth issues? Well, how about 5 GHz 802.11ac channel-bonding routers? I’m getting 70+ Mb/s download speeds from mine with wired connections, and 25 – 30 Mb/s some distance from my 5 GHz wireless link.
Again, looking outside our industry, the two most common signal distribution and switching architectures are based on HD-SDI or IP (or both). Not HDMI, and certainly not HDMI-derived, structured-wire systems like HDBaseT. If the rest of the world wants to multiplex video, audio, metadata, and other low bitrate control signals, they do it over optical fiber. (Did you know that multimode fiber is cheaper than Category 6 wire?)
I’ll wrap up thing by saying that the smart move is for commercial AV integrators to move to an AV-over-IP signal distribution system at the core like everyone else, leaving the HDMI, DisplayPort, superMHL, and “whatever comes next” connections for the far ends, near the displays (if those far-end conversions are even needed at all).
Leave the core as a high-speed, copper bus or optical bus, software-based switcher. If there’s enough bandwidth (and there should be), that system can also carry local TCP/IP traffic, SMTP alerts from connected devices, and control signals to all devices. Not only does this approach free everyone from the “closed world” paradigm of HDMI, it also makes the system infinitely more appealing to end-users and facility administrators, an increasing number of whom come from the IT world.
Consider this ad that was posted recently on a listserv for higher education:
“We are looking for an experienced AV-IT Engineer for the role of Technical Guru. The position will provide planning and support for AV-IT systems used in teaching and learning spaces big and small. The person in this position will focus on design, installation, and troubleshooting of AV-IT systems in a variety of venues, including traditional classrooms, active learning classrooms, large auditoria, computer labs, and even Makerspaces…We are looking for a seasoned professional with a solid background in AV-IT systems. This is a great opportunity for a doer who is excited about not just maintaining but also shaping the future of AV-IT technology as a key element of the teaching mission of one of the world’s top universities.”
I rest my case. It’s time for the commercial AV industry to get in step with the rest of the world and move to AV-over-IP signal distribution.
Wake up. Have you smelled the coffee yet?
Posted by Pete Putman, September 10, 2015 10:53 AM
About Pete PutmanPeter Putman is the president of ROAM Consulting L.L.C. His company provides training, marketing communications, and product testing/development services to manufacturers, dealers, and end-users of displays, display interfaces, and related products.
Pete edits and publishes HDTVexpert.com, a Web blog focused on digital TV, HDTV, and display technologies. He is also a columnist for Pro AV magazine, the leading trade publication for commercial AV systems integrators.