Category: The Front Line

Deconstructing Aereo

On April 22, the Supreme Court will hear arguments in the Aereo case. Broadcasters want the company’s unique “antenna to IP” service shut down, claiming it is avoiding retransmission fees by a clever legal/engineering system, and operates in violation of copyright laws.

Aereo and many other amicus briefs argue that Aereo is within its rights to deliver over-the-air broadcast signals to subscribers because each subscriber has, in effect, their own antenna and cloud DVR, which they control. In this interpretation, Aereo is only leasing space on an antenna to deliver a TV broadcast to subscribers, much the same way an electric company provides power to the home to operate a television.

There have been so many arguments for and against this service that it’s hard to keep up with them. Not only that, two U.S. district courts (1st and 2nd) have ruled in favor of Aereo, barring any injunctions, while a third (10th) has ruled against it, shutting down Aereo’s services in Denver and Salt Lake City.

In an unusual move, the United State Solicitor General has asked to present oral arguments in the case. This is significant because the Supreme Court often defers to the Solicitor General, who (it appears) is preparing to argue in favor of the plaintiffs (broadcasters).

You can find plenty of articles and coverage of the case, most focusing on the legal aspects and current copyright law. I don’t intend to discuss those, but instead will talk about a more obscure aspect of the case – the actual engineered antenna system, and if it actually does what Aereo claims it does.

Aereo Logo 300p

Before I get started, I want to emphasize that I have not seen the Aereo antenna system in person, nor any of the associated electronics. Nor have I had any conversation with company principals. (TechCrunch recently toured the NY Aereo facility and you can see a video of that tour here.) I only have photos of the individual antennas and the antenna array, and the patent applications to go by.

I should also mention my credentials to deliver the analysis you are about to read. I have been playing around with radio equipment since age 5 and actually had two pirate radio stations on the air – one AM, and one FM – while in high school, so many years ago. Not long after that, I studied for and got my first amateur radio license, and within 11 years qualified for the Amateur Extra Class license (KT2B).

Along the way, I spent hours building transmitters, receivers, converters, amplifiers and preamplifiers, and other gadgets, even etching my own circuit boards. At one time, I had a 65-foot antenna tower in my back yard with yagi beams for frequencies from 7 MHz to 2.3 GHz, and have backpacked homemade portable VHF/UHF/microwave stations into the Pocono, Kittatiny, Adirondack, Green, Berkshire, and Catskill mountains.

For about ten years, I wrote a monthly column on VHF/UHF signal theory and operation for the now-defunct 73 Magazine. I also contributed to QST and CQ magazines. When the digital TV transition started, I was the first person in my neighborhood to install rooftop antennas and set-top boxes to watch broadcast HDTV, and of course staged the infamous HDTV Super Bowl parties for ten years.

So I’d like to think I picked up a little knowledge about antennas and signal propagation during the past 45 years. With that in mind, let’s take a look at Aereo’s argument and the flaws I find in it.

THE BASICS – ANTENNA PHYSICS

To receive TV broadcasts, you need some sort of antenna. And that antenna can’t just be a paper clip or coat hanger (although both can work sometimes). The antenna must have some physical relationship to the wavelength of the signal being received. If it does, it approaches resonance and transfers the maximum level of signal to a receiver.

We know the relationship between wavelength and frequency. They’re inversely proportional to each other, and a quick way to determine the wavelength is to divide the frequency into 300. (Or vice-versa to determine the wavelength.) Example: The wavelength of a TV broadcast signal on channel 2, broadcasting at about 55 megahertz (MHz), is about 5.45 meters. If you could actually see the radio wave, one complete cycle of the signal would measure 5.45 meters, or almost 18 feet.

In order for our antenna to resonate – i.e., have gain at the desired frequency – it needs to have some fractional relationship to the wavelength. So, a full-wave loop for channel 2 would measure 18 feet. A  ½-wave dipole would then measure 9 feet, while a ¼-wave whip antenna would measure about 4.5 – about 54 inches.

That’s not to say that our channel 2 antenna wouldn’t work at other frequencies. It could also pull in signals at channel 3, or 4, or even 5 and 6. But it wouldn’t be as efficient at those frequencies as it would on channel 2.

The same principle holds true for high band VHF channels (7-13). To pull in channel 7, broadcasting at about 176 MHz, we’d like to have an antenna with a full wavelength of 5.6 feet. A ½- wave antenna would then measure about 2.8 feet, and a ¼-wave whip antenna would measure about 1.4 feet, or 17 inches.

For UHF TV channels, let’s pick 600 MHz (TV channel 36) for our example. A full wavelength here is 1/2 meter, or about 19 inches. A ½-wave antenna would then be 9.5 inches and a ¼-wave whip antenna, such as you’d find on wireless microphone systems, would measure slightly less than 5 inches.

Again, that’s not to say the ¼-wave or ½-wave antennas mentioned wouldn’t work on higher or lower UHF TV channels. It’s just that they’re most efficient at 600 MHz. All of this is just basic physics and innate knowledge to anyone who has worked with RF antenna and transmission systems, amateur or professional.

Each individual "antenna" in the Aereo system is about the size of a dime.

Each individual “antenna” in the Aereo system is about the size of a dime.

Here is a cross-sectional view of the tiny antenna elements in place.

Here is a cross-sectional view of the tiny antenna elements in place.

Now, let’s look at the Aereo antenna. It’s about the size of a dime and resembles a small loop antenna. Just looking at it in a photo and keeping in mind the science you just read, it would be impossible for such a small antenna to have ANY resonance or gain on low-band VHF TV channels, let alone high-band TV channels.

Yet, that is precisely what Aereo seems to be claiming: One subscriber can activate one of these antennas to watch WABC on channel 7 in New York, or WNET on channel 13. And I don’t see how these tiny little pieces of metal can even work on UHF TV channels: They’re just too small.

Granted, if they were close enough to the transmitting antenna atop the Empire State Building – like a few hundred feet away – the signal levels would be so strong that they would “brute force” their way through the antenna system. But functioning as standalone antennas a few miles away? Not very likely.

Now, here’s where things get tricky and the boundaries between engineering and law become blurred. Aereo installs these tiny antennas in close-spaced arrays on circuit boards. Thanks to the laws of antenna physics, that close spacing guarantees that adjacent antennas interact with each other. That’s due to the principles of inductive and capacitive coupling.

And that means the thousands of smaller, individual antennas couple energy together to act like a larger antenna; one that will approach resonance and have some gain at the desired reception frequencies.

No matter how you switch the antennas, they do interact; it is simple science. And that appears to be the secret sauce behind what Aereo is doing: Creating large “virtual” antenna arrays made up of thousands of tiny, individual antenna elements that, taken together, make up a large, directional antenna array.

According to the patent application, the individual antennas can be switched on the fly to individual receivers, depending on which ones are in use and which aren’t. So the company can claim that each tiny segment of the antenna is actually a stand-alone antenna, assigned to one subscriber.  (Note that, in some earlier Aereo press releases and news stories, they do mention that subscribers can “lease” one or more antennas as needed to pull in a signal. )

THE BASICS – RF, VIDEO, AND MPEG DISTRIBUTION SYSTEMS

Now, if all Aereo was doing was providing thousands of tiny antennas that actually interact to form a large, steerable antenna array, that would be interesting enough. But an Aereo subscription also comes with a “personal” cloud DVR, sitting on a server somewhere on Aereo property.

That means the following must happen for you and me to watch Aereo’s service on our iPhones. (a) A signal must be received from a TV station – say, WABC on channel 7 in New York. (b) That RF signal on channel 7 must then be demodulated by a receiver and converted from the 8VSB modulation format to a baseband video signal, or at least an MPEG2 stream with video, audio, and metadata. (c) The baseband video signal or MPEG2 stream has to be re-encoded or transcoded to MPEG4 H.264 for transport. (d) The H.264 signal is then encapsulated with IP headers and travels to your home network and device.

That takes a lot of hardware. In a conventional master antenna TV system (MATV), one or more antennas are installed on an apartment building or office and one or more amplifiers go with it to distribute the RF signals from the antenna to multiple users. Is this a public performance? From my perspective, no, as the antenna system is merely passing along whatever channels can be received with it. The end-user determines what channels to watch and when. This is a perfect example of a “rented” or “leased” antenna system.

In contrast, a community antenna TV system (CATV, or cable TV) uses large antennas to capture broadcast signals and subsequently demodulates then to baseband video or MPEG, then re-broadcasts them on the same or different channels with a new program guide. In today’s digital world, your cable TV provider has encrypted these local channels, meaning you must lease or buy a compatible set-top box to watch them.

That is indeed a retransmission and a “public performance” in the eyes of copyright law. The CATV company charges for its service and sometimes inserts local ads on those channels. So they provide not only a remote antenna system, they also add in a DVR service, their own program guide, and encryption.

This is why broadcast TV stations and networks have largely given up on the old FCC “must carry” rules and now demand a retransmission fee for their content, just the same way HBO, Showtime, and ESPN do. It’s today’s business model, and it is threatened by what Aereo is doing.

For Aereo to have a 100% true-blue, subscriber-controlled “antenna system,” they would need individual antennas, receiver/decoders, and encoders for every subscriber. That would amount to thousands of discrete pieces of hardware and an enormous capital outlay they’d never hope to recover at $8 per month. Their patent describes a way to assign each antenna to a separate tuner to demodulate the video stream to MPEG2. That might work fine for a handful of viewers. But what if 10,000, 20,000, or 100,000 subscribers are watching at once?

There’s a reason why cable TV companies use single receivers for each channel at their head ends: It’s the only cost-effective way to provide service. And they use multiplexers to route more than one IP video stream to customers for the same reason. It is a classic “one serving many” model and  a cash cow for the likes of Comcast, Time Warner, and Cablevision.

This diagram from the Aereo patent filinhg clearly shows that, at some point, multiple streams of decoded MPEG2 video programs are mixed together and transmitted from the rooftop antenna systems to the MPEG4 transcoders in the basement.

This diagram from the Aereo patent filing clearly shows that, at some point, multiple streams of decoded MPEG2 video programs are mixed together and transmitted from the rooftop antenna systems to the MPEG4 transcoders in the basement.

OK, so let’s buy the argument that Aereo uses a few receivers as needed for each subscriber to pull in TV channels and perform the usual RF-to-video-MPEG conversion. But then, according to their patent application, they combine multiple MPEG2 streams into a multiplex (or “mux”) to send them from the roof of the building to the basement for transcoding to MPEG4 H.264 and ultimately, transmission to each subscriber over an Internet connection.

Combining those MPEG2 streams is really no different than multiplexing TV channels in a piece of coaxial cable delivered to your home. Note that, unlike our MATV example, the TV channels don’t exist in their original 8VSB format. They’ve been converted (altered) to another format for delivery to the viewer.

Note also, in the area between the MPEG-2 Mux and Demux, the words “Antenna Transport (N x 10GBase)”. Here is where Aereo’s entire argument falls apart: You can’t receive an MPEG2 stream with an antenna; only a modulated RF channel. Calling a 10 Gigabit Ethernet connection that streams MPEG2 digital video an “antenna transport” is disingenuous. The signal has to be converted to a new format to travel over this part of the network, and as I just pointed out, it is now a bunch of MPEG2 video programs combined together in one stream for efficiency…just like a CATV or DBS service provider would do at their head end.

In contrast, an MATV system simply receives, amplifies, and distributes RF channels intact to two or more viewers. Those RF signals aren’t demodulated or transcoded – they are delivered in their original state to the viewer. The actual demodulation and decoding happens in each individual TV set.

What’s even stranger is that Aereo is now calling everything ahead of the mux an “antenna.” Horsefeathers! Antennas are antennas; receivers and demodulators are receivers and demodulators. Separate and distinct. That’s as absurd as calling a car an “engine,” or a house a “roof.”

For Aereo to truly provide the service they claim they do, they’d need individual hardware and software processing for every subscriber. No more than one TV channel could travel at the same time to a tuner, and no more than one video program at a time could pass to an encoder, especially not in a multiplexed stream. That improbable and wildly expensive set-up would be a true “leased” antenna and reception system, controlled by the subscriber.

If at any time TV channels, baseband video, or MPEG streams are combined together during the process, then it’s a a CATV system. Pure and simple.

SUMMING UP

Again, let me say that I don’t want to delve into the copyright and business model issues with regards to Aereo. I’ll leave that to the lawyers. Instead, I’m solely focusing on the science of what Aereo does, and to me, it’s overly clever engineering, attempting to re-define the term “antenna” and parse legal terminology.

Their entire argument for getting away with retransmitting broadcast TV content rests on those thousands of individual antennas, which as we’ve learned, unquestionably interact with each other and are separate antennas in name only. The rest of the system appears to be more conventional, with receivers, MPEG streams mixed together, and MPEG transcoding – just like a cable TV does, or even an IPTV multichannel provider, like AT&Ts U-Verse.

The puzzler is why the plaintiffs (TV stations and networks) didn’t pursue this technical angle more aggressively in the first place. In the first court case (2nd Circuit in NY), at least one judge (Denny Chin) called Aereo’s system a “Rube Goldberg” approach, cleverly designed to circumvent copyright law. He hit the nail on the head. There was some testimony from an RF expert at the first hearing, but either the testimony wasn’t presented correctly or contained technical flaws. So the copyright violation angle has been pursued exclusively  by plaintiffs since then.

The judge for the 10th Circuit in Salt Lake City, Dale Kimball, stated in his February decision that “Aereo’s retransmission of plaintiffs’ copyrighted programs is indistinguishable from a cable company.” Kimball got it right as well, as did the three-judge panel that subsequently upheld Kimball’s injunction.

And off we go to the Supreme Court next week. How will this case turn out? No one can say for certain, but the odds appear to favor broadcasters with the Solicitor General seemingly arguing against Aereo.

Stay tuned!

NAB 2014 In The Rear View Mirror

The 2014 NAB Show has come and gone, and although attendance was strong, this year’s edition didn’t have quite the buzz that I expected. Given all that is happening with UHDTV currently, that’s surprising: We are seeing a transformation of television into something very different from traditional models, including demonstrations of next-generation broadcast systems (ATSC 3.0), more powerful encoders (HEVC), and a migration to IP-based video production facilities (the cloud, AVB).

I spent three and a half days at the show, taking it all in while setting aside some time to present a paper at the Broadcast Engineering Conference on the current state of wireless AV connectivity and moderating a Wednesday Super Session on the future of video technology. If I really had to pick one word to characterize this year’s show, it would be “flux.”

Some trends were clear. The Japanese brands (aside from Canon) continue to down-size their booths as their business models shift away from traditional cameras, switchers, recording devices, and monitors. There were numerous companies showing cloud-based storage and media delivery over IT networks, and more than a few booths featured demos of HEVC H.265 encoding and decoding; most of it done with software.

Only a handful of booths emphasized monitors, and some of those had super-sized screens for digital signage out for inspection. In the north and central halls, you could find the traditional purveyors of broadcast transmitters, antennas, and coaxial cable, along with microphones and conventional audio products. But the emphasis seemed to be on “connected” anything – video, audio, cloud storage and delivery, and even wireless cameras for field acquisition and live events.

Given the sheer number of booths, it was difficult to compile a “pick hits” list, but I’ll give it a shot. To me, these companies/products/demos made the trip to Las Vegas worthwhile (and having traveled there over 70 times in the past 20 years, that’s saying a lot!).

Visionary Solutions may not be on your radar, but these clever folks are building some impressive hardware and software codecs at affordable prices. This year, they rolled out their PackeTV system, an end-to-end IP-based video delivery product with scheduling, recording, security, and delivery of real-time and recorded H.264 video, all rolled into one. The graphical user interface (GUI) for controlling the system was well-designed and easy to figure out.

Altera showed how easy it is to stream 12 G HD-SDI (4K) over a single piece of plain coaxial cable.

Altera showed how easy it is to stream 12 G HD-SDI (4K) over a single piece of plain coaxial cable.

 

LG and Gates Air demonstrated their vision for ATSC 3.0, using OFDM and HEVC in a standard TV channel.

LG and Gates Air demonstrated their vision for ATSC 3.0, using OFDM and HEVC in a standard TV channel.

LG and Gates Air had an impressive demo of an ATSC 3.0 concept broadcast. They combined Quad HD, 2K, and SD video programs into one 6 MHz channel, using HEVC encoding and decoding. The signals were encoded at 14, 1.6, and .98 Mb/s respectively, and the signal-to-noise threshold for the SD cast was just 1.5 dB. Multipath sets emulating mobile reception were also demonstrated with the 2K and SD streams holding up very well even at 50 mph.

Sony demonstrated a beautiful 30-inch OLED reference monitor that will soon take its place in the existing Trimaster series. This is a home-grown product and employs the same top-emission system with optical bandpass filters found on the 17-inch and 25-inch products. No price has been announced yet, and Sony has a real challenge in trying to figure out what that price should be as its customers aren’t willing to shell out 1990s bucks anymore for reference displays.

Altera had a clever demo of 12 Gb/s HD-SDI streaming over a piece of “conventional” coaxial cable. 3G HD-SDI has a nominal data cap of 3 Gb/s, so this demo used linked HD-SDI streams to hit the magic number (coincidentally, the data rate for a Quad HD video stream with a 60 Hz refresh rate and 4:2:2 coding). The coax link was 60 feet long and the transmission was flawless, aside from some hiccups on the PC playout server.

Ericsson showed there is more than one way to stream live 4K content. They set up a system that transported a live Quad HD video stream (3840x2160p/60) from a server in England, through satellite and fiber links, to the Ericsson and Intelsat booths at the show. But they used conventional H.264 encoding, breaking the 4K signal into 2K quadrants and using their Simulsync process to stich them together at the receiving end in a seamless presentation on an 84-inch monitor.

Ericsson stiched together four 2K image quadrants to stream this 4K image live from England.

Ericsson stiched together four 2K image quadrants to stream this 4K image live from England.

 

NHK's 4-pound Super Hi-Vision camera records video with 7680x4320 pixels @ 60 Hz and is a marvel.

NHK’s 4-pound Super Hi-Vision camera records video with 7680×4320 pixels @ 60 Hz and is a marvel.

NHK once again had their 8K Super Hi-Vision booth set up, but this time they were streaming live 8K (7680×4320) content from a new, compact 4-pound camera head. The signals were broadcast across the booth in two separate streams on a standard UHF channel, using 4096 QAM at 91 MB/s. Half of the data traveled as a horizontally-polarized signal and the other half as a vertically-polarized signal, both on UHF channel 36. (At that frequency, you can achieve about 20 dB separation between polarization angles.)

Haivision was demonstrating their Secure Reliable Transport (SRT) system over at the Renaissance Hotel. SRT is a hardware/software overlay for existing Haivision encoder/decoder products that is intended to better manage end-to-end streaming over public Internet connections. It offers adaptive streaming rates and two levels of encryption (AES 128-bit and 256-bit). SRT is positioned as an alternative to more expensive satellite backhaul links and dedicated MPLS point-to-point connections.

Korea’s Electronics and Telecommunications Institute (ETRI) had a small but intriguing demo of facial recognition linked to ad servers. The recognition system is built into a standard TV and has a range of about 10 feet, can discriminate between older and younger viewers, and will recognize a face turned 45 degrees to the right or left of center. An appropriate advertisement for the viewer is then displayed during a commercial break.

Fraunhofer HHI always has clever technology demos at NAB, and this year they spotlighted real-time software-based HEVC H.265 encoding and decoding at bit rates up to 40 Mb/s. They also showcased a real-time, hardware-based H.265 decoder using an Altera Stratix V FPGA. This decoder can handle bit rates to 80 MB/s and uses standard interfaces for set-top box designs. Fraunhofer also had an intriguing demo of surround sound playback for tablets in a nearby isolation booth.

BlackMagic's Ursa 4K camera costs only $6,000 (EF lens version)!

BlackMagic’s Ursa 4K camera costs only $6,000 (EF lens version)!

Sony's 30-inch Trimaster OLED looked great, but will its price tag break the bank?

Sony’s 30-inch Trimaster OLED looked great, but will its price tag break the bank?

No plasma? No problem for Panasonic, which showed new 84-inch and 98-inch 4K LCD monitors at the show.

No plasma? No problem for Panasonic, which showed new 84-inch and 98-inch 4K LCD monitors at the show.

BlackMagic Design continues to introduce powerful camera systems at bargain basement prices. Their new Ursa 4K field/studio camera has a huge 10-inch LCD monitor, touchscreen control, RAW and ProRes recorders, and upgradable Super 35mm shutter. The EF lens-compatible version lists at $5,995 while the PL-compatible version is $6,495. Their Studio Camera 4K, also equipped with the 10-inch LCD monitor and 12 GB HD-SDI connections, had an even more amazing price – $2,995.

Intel showed a clever use for Thunderbolt technology: Using a display interface for file exchanges. Thunderbolt runs on the DisplayPort physical layer and has a maximum speed of 20 Gb/s. By using a simple mini or regular DisplayPort cable; two MacBooks, two Windows laptops, or a MacBook/Win laptop can link together for file transfers, working just like a 10GigE network connection.

Panasonic showed it still has game after shutting down plasma manufacturing. Two new large digital 4K LCD displays were up and running in their booth – an 84-inch model (TH-84LQ7OU) and a 98-inch model (TH-98LQ7OU). We’ve seen the 84-inch LG Display LC glass cut before offered by other manufacturers, but the 98-inch hasn’t been in wide circulation. These will replace the 85-inch and 103-inch plasma monitors previously offered.

Finally, Christie had regular screenings to show off its new laser cinema projector system. This projector uses two sets of color primaries and matching eyewear, using wave division multiplexing to achieve a high degree of left eye/ right eye separation. According to a Christie rep, the system can achieve a brightness level of 72,000 lumens, and what was interesting to me was virtually no difference in image brightness through the glasses or without them.

Stuffing Your Brain With Video

A recent poll conducted by the Harris organization revealed that 81% of respondents engage in “binge viewing” on a regular basis – that is, watching two or more episodes of a TV program in a single sitting.

The survey, conducted in mid-March on behalf of Comcast, included over 2,000 adults nationwide and 200 viewers in each of the top ten media markets. Dallas, San Francisco, and Washington D.C. had the highest number of binge viewers among respondents (88%), according to a story on the Home Media web site.

Philadelphia, New York, Los Angeles, and Houston also placed well above 80%, with half of the Los Angeles respondents saying they “binge view” at least once a week. Typically, a viewer decides to check out a new series via pay TV on-demand or streaming from Netflix or Hulu, and settles down in a comfortable chair with food and drink.

I’ve engaged in binging in the past. After CBS began running older episodes of Dexter on late-night TV during the writer’s guild strike a few years back, I got hooked on the show and downloaded Season 2 in SD to my TiVo HD DVR. I followed that with a download of Season 3 in HD, and then began watching on a regular basis via Showtime.

My wife and I would knock off two or three episodes at a time, for that was as long as we could remain seated comfortably. (Dexter episodes, like other premium channel series, usually run about 50 – 52 minutes each without commercials.)

Binge viewing is actually nothing new. The major broadcast TV networks used to run miniseries programming on a regular basis, playing out all episodes of a program during the course of a week. Roots started it all back in 1977, but the difference then was the absence of DVRs – you couldn’t skip the commercials. Miniseries programming ran its course in the 1980s and was largely gone by the end of the 1990s.

To binge view, you need a Netflix, Hulu, iTunes, Google Play, or Amazon Prime account, and an Internet streaming connection (Roku, Apple TV, etc.) or a DVR connected to your pay TV service. And in recent years, we’ve seen DVRs become increasingly powerful: TiVo’s Roamio Plus system has six channels of recording and you can add TiVo Mini satellite terminals to record and watch in different rooms – each Mini takes over one of the DVRs and uses Wi-Fi to stream the program.

Many of us wonder (and rightly so) why we’d want to record six programs at once in the first place. With my circa-2006 TiVo, I can record two shows at once and if need be, use my TV’s antenna to watch a third. But there have been a few times when a third DVR would have been really handy.

Apparently, I’m a piker. Verizon just announced it will roll out a set-top box with the ability to record 12 shows at once, offering enough storage capacity for 200 hours of HD programming. (A good rule of thumb for determining DVR capacity is about 8 – 9 GB per hour for HD programs, so I’m guessing the solid state/hard drives used in the Verizon box, manufactured by Arris, have a maximum capacity of 2 terabytes.)

Memory is cheap. You can pick up 32 GB micro flash cards for about $16 these days and a quick check online shows 256 GB flash drives selling for less that $200 at Amazon. So that 2 TB drive doesn’t add an awful lot to the cost of the new Verizon set-top box. Until Verizon’s announcement, Cablevision customers had bragging rights for the “monster truck” of DVRs, with the ability to record ten channels at once.

Even so, you can pile up programs in a hurry this way, creating a formidable list of time-shifted programs that you may never get to. (We don’t always watch everything we record.) A study conducted by Motorola Mobility (now owned by Arris) one year ago revealed that at least 41% of the programs we record are never watched – yet we continue to schedule recordings and pile up TV shows in our DVRs and complain about not having enough recording space.

All of this begs the question: Why not just stream the programs when you want, and skip the recording process altogether? For binge viewing, this approach seems to make more sense, particularly since you can access a video stream from any platform – TV, phone, computer, or tablet.

The devil in the details is bandwidth. We never seem to have enough of it, and it is costly to expand. During my booth visits at the NAB Show next week, I’ll be paying particular attention to demonstrations of the new HEVC H.265 codec. H.265 promises to slices bit rates by half for any video content, meaning it should be possible to stream 1080p video at data rates in the range of 3 – 4 megabits per second (Mb/s), with 720p streams requiring as little as 1 – 2 Mb/s.

If H.265 really takes off (it’s already supported in some 2014 models of televisions), the balance could be tipped back towards streaming from cloud storage and away from DVRs – that is, if there is a way to retain the commercial-skipping feature that viewers love so much and which you can’t use with most Internet streams.

Perhaps the future model is an online cloud with a monthly subscription that lets you watch shows when you want, anywhere you want, commercial-free. (Oh wait, we’ve got that already – it’s called Netflix…)

Smart Watches Have Not Excited Consumers — But They’re About To

Smart-watch makers are trying to find the right combination of design, functionality, interface, and price that will galvanize a customer population that is curious, but not yet motivated to open their wallets in large numbers. Probably the best-selling smart watch is Samsung’s first-generation Galaxy Gear, which Samsung says sold about a million units through the first quarter of this year despite a $299 price and generally unfavorable reviews. Supply-chain analyst David Barnes (db@bizwitz.com) has told me privately that he is skeptical and believes that if Samsung made a million Gears, at least half of them are sitting in warehouses.

The second-generation Gears (there are versions with and without a camera, and for both the Galaxy moniker has been dropped) are have physical designs that is a bit more refined. Still, they look devices made made by a consumer electronics company, not by a watch company.

As Richard Windsor noted in a recent Radio Free Mobile newsletter, nearly all smart watches look like objects only a nerd could love. What is required for

Motorola Mobility's Moto 360 Smart Watch, here looking like a classically designed upscale watch.  (Photo:  Motorola Mobility)

Motorola Mobility’s Moto 360 Smart Watch, here looking like a classically designed upscale watch. (Photo: Motorola Mobility)

major success is a product that looks like an attractive, upscale watch. Motorola’s Moto 360, which is powered by Google’s new Android Wear OS, could be such a watch. And Motorola Mobility, still under Google’s managerial hand, spent a lot of time and effort on the physical design of the product, including the design of the images displayed. If the price is about right, which I would judge to be $250 or less, this could be a breakout product. Motorola says the watch is coming this summer “in a selection of styles.”

It is encouraging that watch and fashion company Fossil is also working with Google and is likely to be producing watches incorporating Android Wear later this year.

LG and Samsung are also on board with Android Wear, and the LG G smart watch is getting high-level support. “The opportunity to work with Google on LG G Watch was the perfect chance for LG to really pull out all stops in both design and engineering,” said Dr. Jong-seok Park, president and CEO of LG Electronics Mobile Communications Co., in a March 19 press release. Photographs of a prototype show a rectangular, plastic-bodied watch in consumer-electronics-company — rather than watch-maker — style. There’s plenty of time for that to change, though, since LG is saying further details will be released over a period of months.

Moto 360 with turn-by-turn directions on a now smart-watch-looking screen.  (Photo:  Motorola Mobility)

Moto 360 with turn-by-turn directions on a now smart-watch-looking screen. (Photo: Motorola Mobility)

The baseline performance of newly designed smart watches will be established by Android Wear. A highly polished iWatch with a great UI could be a formidable competitor, but in Apple’s post-Steve Jobs era, there is less confidence that the company can pull off anything better than what it’s competitors are already developing. And, despite the endless rumors, Apple has made no iWatch announcements.

If, as is very possible, Android Wear becomes to watches what Windows once was (and to a significant degree still is) to PCs, Google will command the attention of application developers and chipmakers, which will make it even more difficult for others to establish other watch OSs that will take significant market share. In this context, it will be interesting to see if Samsung abandons the Tizen OS for its third Gear generation.

The confined spaces and small battery size inherent in a smart-watch design obviously create interesting problems for the display and display sub-system designers. Motorola Mobility is seeming going with a round full-color LCD. The current generation of smart watches and prototypes use LCD, OLED, and Qualcomm mirasol. E Ink electrophoretic displays have been used successfully in

The LG G Smart Watch prototype, which looks like it was designed by a consumer electronics -- rather than a wristwatch -- company.  (Photo:  LG)

The LG G Smart Watch prototype, which looks like it was designed by a consumer electronics — rather than a wristwatch — company. (Photo: LG)

commercial watches made by Phosphor Watch and other companies, and have also appeared in smart-watch prototypes. Qualcomm’s Pixtronix MEMS

display technology, which I predict will soon eclipse the problematic mirasol, could also be a candidate. But the first commercial Pixtronix-based displays will be in a larger size.

Currently, the most exciting project for me is the Moto 360. For the larger market, I’m very interested in seeing what

Fossil produces. And technically, the LG G might expand the envelope. This is a fascinating new industry, crying its birth cries as we watch.

Ken Werner is Principal of Nutmeg Consultants, specializing in the display industry, manufacturing, technology, and applications. He consults for attorneys, investment analysts, and companies entering or repositioning themselves industries related to displays and the products that use them. You can reach him at kwerner@nutmegconsultants.com.

Samsung Has No Trouble With The Curve

Yesterday, Samsung held its annual home entertainment press event to show off its 2014 line of televisions. Not surprisingly, the emphasis was on curved screens, so Samsung chose an appropriate venue for the event – the striking, spiraling Guggenheim Museum on 88th Street and Fifth Avenue in New York.

At the International CES a few months ago, curved 4K televisions were all the rage. And Samsung wants to lead this category, based on what I saw at the press event. There will be two series of curved 4K LCD televisions in this year’s line – the HU9000 series (55”, 65”, and 78”) and the HU8700 series (55” and 65”).

Prices range from $4,000 to $8,000 for the 9000-series models, while no pricing has been announced yet for the H8700 line.  In addition, Samsung will start shipping the yet-unnumbered 105” curved widescreen LCD (also seen at CES) later this year. And there will be a line of curved 2K sets, numbered H8000 (48”, 55”, and 65”).

Samsung's Ultra HD "tree" it right in the Guggenheim lobby.

Samsung’s Ultra HD “tree” was right at home in the Guggenheim lobby.

There are also some “in your face” models for 2014. The current 85S9 UHD TV (85”) will be joined by the 110S9 (110”). Aside from Vizio’s 120-inch behemoth shown at CES, the 110-inch offering is the largest television ever offered to consumers – and it comes at a dear price of $150,000!

Two of the more interesting products shown at this event weren’t televisions. Samsung’s SEK-2500V UHD Evolution Kit is designed to add some degree of future-proofing to all of the Ultra HD sets in the line. It comes with a Quad Core processor, supports HDCP 2.2 and has HDMI 2.0 interfaces, and is intended to update features of the operating systems on these TVs as needed. (They are, after all, just computers with really big displays.)

Samsung also unveiled a UHD Video Pack (CY-SUC105H), not unlike Sony’s 4K media player. The Video Pack is loaded with five 4K movies, three documentaries, and an assortment of short subjects and costs $300. Supposedly, another UHD Video Pack is on the way later this year.

Samsung's Joe Stinziano and Fox's Mike Dunn discuss the details of the two company's 4K content partnership.

Samsung’s Joe Stinziano and Fox’s Mike Dunn discuss the details of the two company’s 4K content partnership.

A partnership with 20th Century Fox was also announced at the press event to “…establish a secure and sustainable next-generation UHD content ecosystem.” Samsung wants to deliver 4K movies and television programming directly to viewers through its Smart Hub platform (which also got a redesign).

There wasn’t much meat in this announcement, but with television sales continuing to decline worldwide (down by 3% for 2013, according to NPD DisplaySearch) and retail prices for television in free fall, there will be more opportunities for profit in software than in hardware, going forward. And although Samsung continues to sit atop the heap (27% worldwide market share in TV sales revenue), they must be wondering where to go from there.

Conspicuously missing from the Guggenheim show was the company’s 55-inch curved 2K OLED TV. Whether that was intentional or not, Samsung’s focus clearly was on LCD technology and all of the ways it can be manipulated. In the Lewis Theater below the main lobby, Samsung demonstrated its PurColor image processing, along with a new dynamic contrast engine and 4K scaling demos. There was even a side-by-side comparison of 4K and 2K LCD panels displaying standard eye charts to show how much more detail could be resolved in a 4K display.

Members of the press get up close and personal with Samsung's Ultra HD technology demos.

Members of the press get up close and personal with Samsung’s Ultra HD technology demos.

I’ve written previously about the HDMI 2.0 interface and how it is barely fast enough to support 4K RGB images with 8-bit rendering, and have wondered on more than on occasion why more TV manufacturers don’t incorporate the faster, royalty-free DisplayPort interface. Well, I spotted one on the back panel of the Samsung Ultra HD TV used for the PurColor demo, although it was marked “TV One Connect” and no one seemed to know much about its function. (A subsequent follow-up with a Samsung executive clarified that this port is designated for an “evolution” hardware/software connection.)

Almost lost in all of the hype was the fact that Samsung will have a pair of 75-inch flat screen 2K LCD TVs in the lineup for this year. The H7150 version will carry a price tag of $3,999, while the “entry level” H6350 will be tagged at $3,299. These prices are competitive with Sharp’s 70-inch and 80-inch Aquos 2K sets that are wildly popular for commercial and educational installations, and I suspect Samsung wants to grab a piece of that action for itself.

They jury’s still out on the benefits of curved screens. Yes, Samsung did talk about everyone having a great viewing angle and how curved screens make viewing a more immersive experience. That’s certainly true for the 105-inch 21:9 screen, but I don’t see the advantage for 55-inch 16:9 TVs. And if there are any sources of glare, the curved screen doesn’t eliminate them – it just moves them to some other viewer’s disadvantage.

I will admit, though, that curved TVs do look cooler than flat screen sets. And the choice of the curved, soaring Guggenheim (designed by Frank Lloyd Wright and opened in 1959) as a showcase was inspired.

Now, let’s see if consumers are similarly inspired to purchase a curved TV…