Wednesday, November 27, 2013

The Other Side of the Bridge - A Look at Some AV Streaming Solutions

One comment I frequently hear about HDBaseT is that it's a "bridge technology" between the old days of simple point-to-point connectivity and a future in which AV joins the rest of our data on those great big IP networks which dominate the rest of our lives. Are we ready for this paradigm shift? Is the future here? Perhaps not, but it's tantalisingly close. I've recently had a chance to look at two compressed video over IP products : Just Add Power's 2G+ system and SVSI's N2000 series. While neither seems quite ready to dethrone HDBaseT as the defacto video standard, we've reached a point at which these types of solutions deserve, at the very least, to be part of the discussion.
The innards of a Just Add Power
endpoint

Form Factor and Convenience
Form factors for the base transmit/receive units are similar, dominated by the familiar six-inch square flattish metal box Just Add Power adds a three-encoder rackmount unit while SVSI offers a card-cage for flexible configuration of permanent installations. SVSI's standalone units are UL rated for use in plenum spaces, allowing the to be safely (and  legally!) installed above ceilings in most localities. What's more, PoE (power over ethernet) is standard with SVSI and an option from Just Add Power. This means that one really need run only a single cable. 

Performance
The units all performed as advertised, albeit with their own quirks. The Just Add Power demo kit, consisting of transmitters, receivers, a network switch, and a wireless access point, was delivered to me pre-configured with each switchport configured for a particular device. That's right, their configuration apparently requires you to know which device is going to which port and to configure the switch accordingly. Once I got the wiring straight it worked as advertised; switching was quick, and the system boasted a handy "video wall" mode in which it would tile an image across four or more displays without any extra hardware. The switch configuration issue is a bit of a concern to me; this will need to change, but at present AV installation techs don't tend to be the best at IT configuration. In fact, one often gets a blank glassy stare sometimes after "is it turned on" and "are all of the wires plugged in". A look at the manual for their software seems to indicate that switching is handled by putting switchports on unique VLANs and moving these around to match the VLANs of the destination. This strikes me as an odd way of using a switch, but I'm not a network engineer by any means.

SVSI's N2000 units each had the now-familiar web interface, showing stream ID numbers, scaling, audio embedding, HDCP status, etc. It also handles switching and routing a bit differently than Just Add Power; as is the case with the N1000, every encoder is assigned a "stream number". A decoder can then choose which stream to receive. There is also a multicast option for greater network efficiency.  Other controls include a slider for video quality, selection of scalers, image cropping, and HDCP enable/disable. Image quality is quite good, but at the expense of noticeable lag. Such is the price of video compression. It isn't enough to make it unusable by any means, but would be an impediment to realtime collaboration or annotative applications; if one sketches something in a drawing program one doesn't want the line on the screen to trail the real-time activity.

SVSI's units have onboard scalers, which are a nice tough although somewhat limited in what resolutions they can handle. A test monitor with a really weird native resolution ended up with severe underscan, while more standard 1920x1080 displays worked perfectly well with a variety of inputs. On the positive side, the web interface gives the full extended EDID for those who need to know exactly why their image doesn't look the way it should.
 

Tiling and Windowing
Not only could images be strategically cropped to create a tile-effect (as above), but they gave me an additional toy with which to play - a 4-input windowing processor. The inputs in this case are streams from N2000 or 1000 series encoders, and the web interface allows one to create layouts of up to four windows. With one of these windowing processors per display and a bit of creative cropping one can build a complete video-wall of pretty much any configuration so long as no more than four windows touch any single element. Is this quite as flexible as other forms of window processing, but is more than adequate for some applications. It's another case in which the video over IP technology is catching up to everything else.

I'll aside here that Christie has also made a move into the IP world with their Phoenix system; Phoenix endpoints are connected via IP and can send to and receive from each other through a standard gigabit ethernet switch.O It's an interesting product in its own right, likely deserving of its own post. For now, we can take it as another sign of how things are moving; solutions which a few years ago would have required dedicated copper or even fiber ones can now be part of the same network as the rest of ones data.

Building an AV Ecosystem
One of the more exciting things about having AV on a network is the possibility of creating a unified ecosystem, in which live content, signage, and a larger unified communications platform all work together. Software players exist to bring H.264 and even JPG2000 content to PCs (although the latter might be somewhat restricted in frame rate if you don't have a fast enough machine). A plethora of recording and processing options are available for digital content. IP based systems can, at their best, change the way we look at an AV installation from individual systems to an interconnected AV ecosystem in which various resources can be called upon not only in various conference rooms, but also on desktops, tablets, and sent to remote locations. 

We started by asking if HDBaseT is a bridge technology. I'll close with a different, and more interesting question: how does replacing HDBaseT with network transport fundamentally change what an AV system is and how we interact with it?

Wednesday, November 20, 2013

The Ones Who Walked Away from Denver

And now for something completely different. The hint of a literary reference in the title makes this one, broadly speaking, an "ink" post. What we're really talking about today is ethics.

The one who walked away from Denver - the person whose story inspired this post - is a now-former professional football player named John Moffit. After a couple of years at the periphery of the National Football League, Moffit made the choice to walk away while he still could walk, turning his back on what many of us would see as a considerable sum of money. He made this choice after an offseason spent reading the works of thinkers including Noam Chomsky and the Dalai Lama, after reflecting, and after concluding that the expectations for his life were part of a larger machine disinterested in his personal well being. Not that many years ago I'd have tipped my hat to his personal choice while lamenting the loss of football talent for the Autumn Sundays' spectacles. Today, I just tip my hat, as I've walked away from Denver myself a long time ago.
John Moffit

The title of this post, for those who don't read the same kinds of stories I do, is  nod to Ursula K Leguin's short story "The Ones Who Walk Away From Omelas". In a nifty coincidence, I just learned that it was published four decades ago last month (and been anthologized approximately 17 million times since then). Those who haven't read it yet can feel free to do so now.  Go ahead. I'll wait. (Aside: those of you who don't know Le Guin have some marvelous reading ahead of you!
Ursula K Le Guin. A national treasure.
She's a speculative fiction author, but also one of the most brilliant and talented American fiction writers of the latter half of last century. There are not words for me to greatly enough recommend her work. That should, perhaps, be a separate post. End of aside.)



Are you back yet?

Great. To recap those those of you who didn't take a short-fiction detour, Omelas is a lovely little utopian city which owes its success to the suffering of a small child. Everyone knows this, and most go on with their lives. Some, however, on learning this, choose to walk way and leave the comforts and wonders of Omelas behind them. One lesson I take from this (with the caveat that boiling any piece of literature down to a single lesson is reductive in the worst way possible) is that there is value in the choice to not take pleasure from the suffering of another, even  if your walking away does nothing to reduce that suffering. You can speak it aloud, you can make yourself an example or, at the very least, feel that you are acting within your values system.

My argument against football, for those interested, is that the game itself is inherently violent. Serious injuries, especially repeated traumatic brain injury, are so much a basic part of the sport that I don't think they can be redressed significantly enough for me to be able to watch in good conscience unless the game itself were fundamentally altered. There's also a celebration of violence within football culture; we applaud the dramatic hit, the hard tackle. I will (and do) watch baseball in which injuries still happen but are tangential to the actual game, not the apparent goal of it. Your mileage, of course, may vary but it is something I urge you to consider thoughtfully. Choose to watch or choose not to, choose to celebrate or choose not to, but do so mindfully knowing that this, like all other things, is an ethical choice.

Circling back to my professional life, I'm fortunate to work in a field which is, for the most part, not ethically problematic for me.  Many of the clients I've worked for are in the higher education, government, or healthcare verticals. What would I do if we won a football stadium project and I was handed a part of it? That's an interesting question, and one for which I don't have a ready answer. Would I risk being labelled as less than a team player by turning the work down, requesting that someone else work on it, or even resigning were my hand forced? Would I decide that the issues with football - including the political and economic issue of public funding of for-profit businesses - are great enough to risk my career? That I doubt. It is something, however, I would approach mindfully and make careful judgments as to how I can balance 

Wednesday, November 6, 2013

HDBaseT Interoperability Follies

Welcome back to any AV followers who faded away last month while I was posting all fiction all the time. The stories and poems won't entirely go away but we'll start moving into a little better balance between pixels and ink for the nonce. Today I'd like to share a moment to discuss everyone's  favorite transport mechanism, HDBaseT.

For those not in the know, the HDBaseT alliance have defined HDBaseT as a technology for transport of video, audio, power, ethernet, and control over standard category cable. As a point-to-point (as opposed to routable) signals, HDBaseT cannot be routed with standard network switches. With offerings branded by major techm manufacturers (ie, Crestron's "Digital Media" and AMX "Enova" it has become a defacto standard for commercial systems.

Or has it?

The idea of a "standard" is that it should be manufacturer-agnostic. The HDMI output on a Sony Blu-ray player, for example, will work on the input of a Sharp LCD panel as well as it would on a Sony. Is that the case with HDBaseT? Is it, in other words, really a "standard"? The HDBaseT website does boast, in a large banner, that it is "The standard of the future", but the remainder of their text is much more cautious, referring to HDBaseT "technology" or "specifications". Can one grab an HDBaseT transmitter from one manufacturer and receiver from another and expect them  to work together? In an ideal world, the answer would be yes. Those of us who have wrestled with HDCP, EDID, or other HDMI-inspired headaches knows one thing for sure: this is not an ideal world.

For the sake of my own curiosity I tried a little experiment. I had access to transmitters from three major manufacturers: Crestron Digital Media, AMX DXLink, and Extron XTP. The latter is not, to the best of my knowledge, HDBaseT certified, so I'd not expect interoperability from it. The others are, so I would. Is that what happened? Not quite.

The AMX receiver gave me a "green screen of doom" when I tried to connect the Crestron or Extron transmitter to it. This is apparently an HDCP handshake error (as opposed to HDCP noncompliance, which would give a red screen of doom. Doom is the consistent part). This possibly has to do with how AMX handles HDCP authentication, which treats the receiver as a source rather than a repeater. This is nifty in that it bypasses the key limit in some sources (in other words you can run a single Blu-ray player to as many displays as you have outputs for, regardless of its key limit), but might make the receiver more picky in what it looks for from the source side.

This was a fairly disappointing result in that it left the non-HDBaseT certified Extron XTP as my only other receiver. IT might not be "certified", but it does use the same technology as HDBaseT transport solutions. Somewhat to my surprise the Crestron and AMX transmitters both sent HDCP protected content to the Extron XTP receiver. So much for certification.

What does this mean in practical terms? Not all that much. What it highlights is just how similar these devices are. Even in terms of form-factor you get a great deal of similarity between product lines; everyone has a standalone receiver about an inch or so deep to fit behind a flatpanel (or in a wall box), a similarly shaped standalone transmitter, a two-gang wall-mount transmitter,  and modular matrix frames sized from 8x8 up to 32x32 or larger. There's no real practical reason to step out of a single manufacturer's ecosystem other than to prove that you can. Now that we've done so, even that is gone.
A selection of transmitters and receivers

So how does one choose? We're the same place we were back when we did "Switcher-Wars" last year. Do you need Crestron's full audio and USB breakaway? AMX's smaller form-factor and better energy efficiency? Does an end-user with limited programming expertise want to be able to make equipment substitutions and other programming changes through Extron's Global Configurator or AMX's Rapid Project Maker? Is there an existing implementation of a remote-control and asset management system (Crestron Fusion, AMX RMS, Extron Global Viewer) into which you'd like your new system to tie? We've reached a point at which not only are the differences more subtle, but many of them won't even appear on a spec sheet.

The real take away here - for those who didn't realize it already - is confirmation  that the technology is very much the same. In fact, it's the same to the point that if one files off the serial numbers one would have a hard time even telling one apart from another. What this really means is that the best manufacturers aren't just selling the technology; they're selling a solution, including an ecosystem to fit around it and the thought they put in to some of the details that one might miss on a spec sheet.