4K Smartphones - Stand and Deliver

In my last post I ventured to predict that 4K displays will become mainstream on high-end smartphones at some point even though they’re unlikely to deliver a better viewing experience for the mobile user on average. To become a game changer, it takes more than just adding a impressive new feature to a device to attain that immersive cinematic experience. What about 4K video content? Is enough readily available? Are the wireless networks and mobile infrastructure in place to transport high-resolution content? What are the effects of 4K displays on the smartphone itself? It’s a tall order to provide a comprehensive picture in a blog, so I’ll attempt providing some vignettes of insight in a question and answer style. We’ll shortly probe the ecosystem from production through to transfer and finally consumption of 4K content to arrive at the bigger picture.

"Looks pretty much like 4K to me"

Seeing the difference

Spicing up the quality of images or video is not only a case of increasing resolution. It’s technical and personal matter at the same time.

What parameters influence the perceived quality of a display?

A combination of technical factors such as the screen size, display resolution, frame rate, color depth, dynamic range, viewing distance and other parameters make up for the quality of the viewing experience. In addition, your own eyesight plays an important role, in particular its ability to distinquish fine details, known as visual acuity. The resulting mix of objective and subjective factors makes it extremely difficult to pinpoint a single metric as being paramount in providing a superior viewing experience.

What does it take to produce 4K broadcast quality content?

Producing 4K content dictates a four times higher resolution than the prevalent HD (high-definition) standard, resulting in 4x the amount of data too. That translates to very expensive equipment in the production chain (cameras, lenses, switchers, encoders, storage disks, editing workstations) so that only a trickle of material is currently shot in 4K. Over time equipment pricing will erode and options will increase. Shooting in 4K will ultimately ensure that content is future proof, spurring on its adoption.

What content is available in 4K resolution to date?

YouTube Netflix, Amazon, Sony, DirectTV, Comcast are just some of the names who boast 4K UHD movies, TV shows, or video clips, all of which have caveats of sorts. A good overview of 4K UHD programming available in the USA currently (December 2014) can be found at Digital Trends. Indeed, it’s quite a limited offering but it’s an incoming tide that’s rising.

Squeezing 4K data

Getting 4K content to the end-user requires high-bandwidth transmission paths. 4K codecs that efficiently compress the material when it is captured (encoder) and decompress it again when viewed (decoder) play a pivotal role in easing bandwidth requirements and the cost of transfer.

What are the transport requirements for 4K video?

As can be seen from the listing of offerings at Digital Trends, most 4K content is streamed over the Internet and requires a 25 Mbps channel. Many experts contend that 36 Mbps is the best channel bandwidth for delivering Ultra HD, whilst others purport that 15 Mbps is sufficient for a decent 4K experience that is also commercially viable. Whatever the case, that’s certainly a lot more compared to delivering a HD (1K) stream at 3 Mbps. It also costs approximately 5x more.

	SD (720 x 576)	HD (1920 x 1080)	UHD (3840 x 2160)
Recommended Internet Speed	3 Mbps	5 Mbps	25 Mbps

Netflix’s recommendations on Internet speed for viewing their movies and TV shows

Why are differing minimum 4K bandwidth requirements quoted by experts?

This depends on the type of content and the codec used for compressing the video footage. For example live sports events with fast moving action will require higher frame rates (more data to encode) for a fluid viewing experience. In addition the encoder used to reduce the data must compress at a faster rate due the live nature of the event. Stated in simple terms, the more time a codec has available, the more it can squeeze the data without sacrificing quality. So it stands to reason that films shot and compressed in the studio require less transport bandwidth than live sports events broadcast on the fly.

What codecs are available for 4K video compression and do they differ in the quality and bit-rate?

Of late there’s been a pitting of solutions between Google’s VP9 that follows an open source approach, and MPEG / ITU groups’ HEVC (High Efficiency Video Codec, also known as H.265). The latter’s previous H.264 (MPEG) codec has been the defacto standard for HD content in the past, but Google is hot on the heels to change this for 4K. Then there’s another initiative from Mozilla and Xiph.org called Daala that claims it will beat both on technical merit. In a future post I’ll attempt to uncover some of the key differences of these codecs in their complexity, delivered quality, achievable bit-rate reduction, available content offerings and devices that support each.

Please, no frame freezes or fallouts

Many of us know how annoying a sudden and unexpected throttling of bandwidth can be, especially when enjoying streamed video or audio. Consistent channel bandwidth is of the essence, so let’s take a look at the options.

What channels are available for transmitting 4K Ultra-HD content?

Several options are available such as satellite, microwave, cellular and fixed networks using cable or DSL. Currently the majority of 4K content is available as video-on-demand (VoD) that is stored and then streamed through fixed networks (cable/DSL) to the end user. On the broadcast TV front, satellite providers like DirectTV and BSkyB will lead the 4K race. Terrestial TV broadcasting will require more time for standards to come into place and availability will largely depend on country initiatives in transitioning to newer technology with Japan and Korea at the forefront.

Will mobile networks have sufficient bandwidth to transmit 4K content?
4G cellular networks typically support download rates of up to 150 Mbps for LTE Cat 5 smartphones that are prevalent on the market . More than enough for one UHD channel at 25 Mbps one would think. Keep in mind though, the cellular broadband network is being shared by many people at the same time. In fact we’re dealing with a top theoretical download speed that most of us won’t ever witness in the real world. Even worse, the data rate may change unexpectedly depending on network usage, and your 4K video may suddenly freeze.

Can Wi-Fi reliably stream 4K video content?

The more recent WiFi standards that employ 40 MHz channel bands (802.11n or 802.11ac) provide sufficient data throughput to support 4K video transmission in theory. In practice however, performance is unpredictable as those of us who have tried streaming HD video at home will know. This is largely due to contention between neighbouring WiFi networks whose data rates start sagging as they counter the chatter of next-door access points on the same frequencies. In addition WiFi microwaves, especially those operating in the 5 MHz band, are attenuated by walls and floors, leading to a 80 - 90 % drop in peak rate compared to the access point being in the same room. In short, uninterrupted video delivery is simply not reliable enough.

What other wireless standards can handle 4K streaming and when will they become available?

Two new standards, WiGig and WirelessHD, are out there and both operate in the 60 GHz band. They are much faster than Wi-Fi 802.11ac or LTE mobile broadband technologies. Their formidable throughput rates (7.5 Gbps for WiGig) are targeted at the wireless delivery of high-definition content. Yet the 60 GHz signals they use cannot penetrate walls. It’s all about connecting computing and entertainment devices in the same room without cables. The idea is to turn mobile devices into media stations that wirelessly dispatch streams to 4K TVs and displays. Devices supporting the new standards are expected to start shipping this year and they will allow uninterrupted wireless streaming of 4K content, albeit in the same room.

Expensive power hogs

Smartphone innovation happens at a mindboggling rate and users often wait for announced models to be released before replacement. Even if 4K is one of those desirable new features, there are other factors to observe.

What are the drawbacks of smartphones with 4K displays?

Power and cost are two inhibitors in the uptake of 4K displays on smartphones. The display is one of the major culprits in draining the battery on a mobile phone and the amount of power consumed is directly related to the size of the display. 4K also means four times more data to store and process. More memory and busier units on the SoC (system-on-chip) will sap precious juice from your phone’s battery even faster. In addition, the display is the component that costs the most in a smartphone. Paying more and having to charge your phone more often don’t speak for a runaway market success.

Can smartphones record 4K video?

The minimum requirement for a smartphone camera to capture 4K content is a camera with at least 8 megapixels (MP) of resolution. Most mid-range devices already sport such pixel magic. But can available smartphones capture and encode 4K video? Suprisingly yes, even though most don’t feature 4K displays. For example, both Apple’s iPhone 6 and 6 Plus models (8 MP camera) support 4K video recording; Samsung’s Galaxy S5 (16 MP camera) and Sony’s Xperia Z2 (20.7 MP camera) both use Qualcomm’s Snapdragon 801 that supports video capture and playback using the HEVC / H.265 codec. In fact, an impressive list of smartphones supporting HEVC can be found at phoneArena.com which are able to record 4K videos.

A nascent technology gaining momentum

What do you do if 4K content is limited, seamless transfer paths still patchy, the first smartphones with 4K displays just around corner, yet many featuring 4K capture available right now? 4K’s marketing machine has the answer and is harping that user-generated 4K content on mobiles will bridge the gap. Smartphones will be the force to bring 4K into the limelight as users record 4K footage and either watch it on 4K TVs or upload it to YouTube. A bulletproof business model or high hopes on an act of faith? Whatever the case, it is apparent that the industry is working its fingers to the bone towards a single new display standard. At some point the law of large numbers will bring down costs and herald in the age of 4K phablets, one way or another, sooner or later. The market research company ABI Research forecasts almost 500 million 4K display-enabled mobile phones to be sold in 2019.

It’s still an HD world out there and will be for quite some time, but 4K UHD is on the move.

Spectrum - private or public beachfronts?

Spectrum seems similar to real estate. At least that's the image that is conjured up by the often used analogy "public or private beachfront?".

You could argue the beach (spectrum) belongs to government (regulator) who should plan its best use for its people, either by selling it off (licensed spectrum) to an investor (carrier) with strict clauses for its development (cellular infrastructure coverage) and ensuing use by paying visitors (subscribers), or by keeping it a public space open and free to all (unlicensed spectrum). This seems like a pretty good analogy but it seems to ignore one key contributor that is throwing a spanner in the works: technology advancement, or more accurately, semiconductor improvements. Wireless data is exploding through the usage of new wireless handhelds and devices and there is no end in sight. The Internet-of-Things is looming, further intensifying the situation as seers predict everything with a current running through might potentially be a source of further wireless data. So all sights are set on "spectrum" to offer a cure, and governments (regulators) who ultimately own it. Indeed, we need more spectrum. And lots of it.

Spectrum is big money

Spectrum is where big money meets legislation meets high-tech, yet always bound to the perennial laws of physics. It's a billion-dollar industry that encompasses auctioning, immense investments for operators who build nation-wide networks, influenced by a constant stream of new technologies that improve connectivity for all. Proponents of the one camp (licensed) rightfully argue that the immense investments for carriers building nation-wide networks require long-term perspectives for ensuring a profitable business. The other side (unlicensed) maintain that only open policies guarantee fair and efficient use of the airwaves and lead to universal, affordable and ubiquitous broadband for all. A case of "Cellular" vs. "WiFi" is in the ring for the next round.

"Spectrum Crunch" - a myth?

The crux of the issue is that much of our spectrum, particularly around the "sweet spot" is already in use. At least that what the skeptics purport, creating a shortage or "spectrum crunch" myth. The truth is a little more varied and subtle though.

Click to enlarge

Radio spectrum extends for 3 kHz to 300 GHz and mobile phones currently operate between 700 MHz to 2.6 GHz in this range - the so-called "sweet spot" based on state-of-the-art technology. Note, that this sweet-spot spectrum block represents roughly just under 1 % of available "radio spectrum". Spectrum shortage? Well, yes, in a way, if  you regard what's exploitable using today's available technology. Just as broadcast AM radio in the early decades of the last century relied on valves and tubes for frequencies up to 15 kHz, today's semiconductor technology defines wireless communications around the 1 GHz mark. But what about tomorrow? 

First approach: making the best of "redundant" spectrum

To date specific blocks in the full radio spectrum range have been allocated for specific purposes or services by governments, and most of these are historical in context. For example blocks for maritime navigation, aeronautical navigation, public safety, broadcast TV etc. A true patchwork as can be seen in Sebastian Anthony's "wireless crunch" blog for ExtremeTech. Of course spectrum is a unique and therefore valuable resource, but it does differ from oil, gas, minerals etc. which face continuous depletion until none are left. Spectrum can be revitalized by closing older, inefficient services, as has or is being done for analog TV, or by making use of "white space" consisting either of unused spectrum blocks or guard bands between used blocks/frequencies that were required in prior times for broadcast and communications technology to work.

Technology advancements increase spectrum usage

As the needle of technology heads towards the future, better usage in existing spectrum blocks and new usage in regions higher on the spectrum map become available accordingly. Ever decreasing semiconductor process nodes ensure the implementation of mathematical algorithms providing better modulation techniques and other advancements that ultimately increase bandwidth and bit rate per Hertz of frequency. Of course at the heart of cellular and wireless communications is the trade-off in providing the best capacity (more bandwidth/higher bit rates using higher frequencies) for the largest possible coverage area (lower frequencies are better here). Broadly speaking, lower frequencies travel further than higher ones but cannot carry as much information. The current sweet spot for wireless communications is determined by these facts and primarily governed by what's technically possible based on state-of-the-art semiconductors. Advances in semiconductor technology will allow the sweet spot to travel further up the spectrum over time and provide better bandwidth. The diagram below shows how these advancements will ultimately lead to the still amorphous 5G technology as we get closer to the year 2020 where wireless Gigabit speeds will become reality.

Click to enlarge

Does your devices' antenna size still matter?

Fitting a large antenna or satellite dish on your roof works for great TV reception, but what about handheld/portable wireless communication devices? One of the limiting factors of wireless devices has always been the required size of the antenna for transmission and reception. Antennas "resonate" at one quarter of a frequency's wavelength. Taking this factor into account for the typical current "sweet spot" frequency at 1 GHz leads to a minimum necessary antenna size of 7.5 cm which resembles an acceptable cell phone size. At 300 MHz this increases to 25 cm whereas at 3 GHz we get down to 2.5 cm. In other words, higher frequencies spell out smaller device form factors, or more room for integrating antenna arrays, read MIMO et al. As these new antenna technologies come into play, data rates can be increased through better wireless performance in several respects, irrespective of the maximum bit rate achievable by the modem dictated by modulation techniques/Shannon's theorem.

Some worthwhile spectrum vs. technology resources

Spectrum, it's allocation, management and better exploitation through technology is a complex field in itself. Both free and subscription-based resources provide great insight into the issues at hand. Regarding today's 4G technology, we've reached the max in terms of squeezing as many bits as possible per Hertz of spectrum, as eloquently reported by Kevin Fitchard at Gigaom on wireless networks as workhorses of the web. A further great reference is Frank Rayal's blog looking at the capacity-coverage issue for licensed and non-licensed band. It investigates and compares differences in this respect for LTE and WiFi, both using a 20 MHz band. Certainly both camps have convincing points to advocate as we move further down the timeline. Telecoms and spectrum expert Gunjan Idrayan writes a regular insightful blog on spectrum and wireless with a particular focus on the situation in the United States and India. You'll also find some timeless articles about licensed and unlicensed spectrum as well as mobile broadband written by Peter Rysavy. Furthermore here's a short list of paid research that is well worth investigating if spectrum belongs to the core of your professional work.

The Global Spectrum Database PolicyTracker	http://bit.ly/THmNBb
Spectrum Auction Tracker Analysys Mason	http://bit.ly/1t5EQkV
2G, 3G, & 4G Mobile Network Subscriptions, Spectrum Licensing, Ownership, Infrastructure Contracts & Handset Shipments Database Signals and Systems Telecom	http://bit.ly/1lEZ63o

Spectrum for Mobile Broadband IDATE	http://bit.ly/1lEZGhx
Radio Spectrum Intelligence Tolaga Research	http://bit.ly/1mR4sc9

There are many more market research reports on the topic of spectrum available, and you can also stay on top of the latest worldwide events by signing up at wi360 (free). You will receive email alerts on new research and events regarding the specific topic of spectrum.

Spectrum's future

Governments that regulate spectrum and the technology sector that provides the devices for wireless access are both doing their best by squeezing the most out of a limited resource. Yet each of them works with a completely different heartbeat and thus the "divide" is increasing at an alarming level to the detriment of the consumer and subscriber. Wi-Fi over the past decade is a case in point how unlicensed spectrum has been addressed remarkably well at breakneck speed outside of government intervention. At the same time, network investments in the licensed spectrum must also remain a lucrative business for carriers. With the subscriber in mind, the golden age of raking in exorbitant cash amounts by governments through spectrum auctioning requires a future rethink.

Our spectrum future will be mapped out how well concepts from all parties will be mutually respected and implemented.