Search This Blog

Tuesday 28 June 2016

FLAT lens in the broadcast technology

A new report in the journal Science points the way to a new lens of the future that could revolutionise optics entirely.
Lenses are curved, we all know that: that’s how they bend and focus light and that’s how they’ve done it since around the time of Aristophanes who wrote of a ‘burning glass’ that was used to light fires in his play The Clouds around 2500 years ago.
As we’ve seen recently with advances in the fields of computational photography and the like though, just because a thing is one way it doesn’t mean it necessarily always has to be so. And a new paper in the US journal Science, ‘Metalenses at visible wavelengths: Diffraction-limited focusing and subwavelength resolution imaging’ points the way to a very different future indeed.
Metalenses are flat. Very flat. And possibly cheap to produce too.
As reported by the BBC, the lens developed by the team behind the paper is made of a thin layer of transparent quartz — paint whitener, or titanium dioxide to be exact — which effectively coast a sliver of planar glass in millions of tiny pillars, each just tens of nanometres across and hundreds high. Individually, each of these pillars reacts with incoming light, their combined effect being to “slice up a light beam and remould it as the rays pass through the array.”


At the moment we’re talking very much about the micro scale — the prototypes produced so far have been limited to 2mm across due to the constraints of the manufacturing equipment at Harvard University where the research has been carried out. But so far the new optics have performed rather astonishingly; outclassing top end lenses used in research microscopes by removing the aberrations inherent in the glass and producing an image rated as 30% sharper.
Of course it is very early days, but the technology here could truly be a game-changer for the optical industry. The new metalenses could be made in the same fabrication plants as computer chips for a start, and with mass production also comes the potential to scale up the size.
"Once you have the foundry - you want a 12-inch lens? Feel free, you can make a 12-inch lens. There's no limit,” comments Professor Federico Capasso of Harvard University, the senior author of the report.
Cell phones are the first target of the tech, with Prof Capasso also reeling off a list of other uses such as quality control in factories, light-weight optics for virtual-reality headsets, even contact lenses (as the metalenses can be made with soft materials). Production lenses? We’re still at the first prototype stages, so it’s not quite time to stick that case of Primes on eBay just yet. But we are getting enough inputs now to suspect that a tipping point may not be too far in the future. The future decades of image capture could be very different from the ones that have gone before.

UHD

“Ultra HD,” of course, means much more than just higher spatial resolution (4K and 8K), even though that is to what many in the industry and consumers equate it. It includes four other image as well as immersive audio technologies. The four other image technologies are high dynamic range (HDR), wide color gamut (WCG), deeper sample bit depth (higher number of coefficient bits for the digital samples), and high frame rate (HFR). All of these technologies combined create a much more realistic and immersive TV viewing experience than today’s conventional HDTV, without requiring any special head gear (such as that required for the near-totally immersive virtual reality techniques that are also a new exciting area of discussion). 
BANDWIDTH CONSTRAINTS
As is typical, pre-produced content is ahead of live TV broadcasting for delivering UHD, with Ultra HD Blu-ray discs and some on-demand or over-the-top services already providing both 4K and HDR. This should not be a surprise: it’s much easier to implement new services when one controls both ends of the ecosystem, with only a passive delivery pipe in between. All post-production is done in advance, with the receiver/player/rendering/etc. acting on the content, unchanged by the delivery mechanism, whether via wire, wireless or “sneaker net.” With live TV broadcasting, there is no advanced post-production; all is done in real-time. Live changes (such as logo insertion, lower thirds, graphic overlays, picture-in-graphics, squeeze & tease, cross-fades for interstitials, or country-specific format conversions, etc.) could occur at each “node” along the live broadcast chain. Another major factor is available bandwidth; there simply may not be sufficient bandwidth to delivery UHD services, especially for the technologies that require a lot more bandwidth than today’s conventional HD services do. Lastly, there may be regulatory restrictions impacting new services over certain networks, such as over-the-air broadcasting, that prevent economically-viable transmission of UHD services.
As previously mentioned, all of the five image technologies combine to give a much more compelling visual user experience than today’s conventional HDTV, so if a content/service provider is able to offer a UHD service that includes all five, then fantastic. Definitely offer this compelling, much more realistic TV viewing experience. Consumers will be thrilled.
However, what if bandwidth constraints prevent the content/service provider from offering “full UHD,” so to speak? Should the provider just resign itself to continue with today’s conventional HD service? Absolutely not. The combination of HDR + WCG + 10-bit sample depth (which many of us refer to as “HDR+”) has been shown in numerous demonstrations to have fantastic “wow” factor for consumers. And the good news is, depending upon the HDR+ scheme chosen, the bandwidth increase over conventional TV (what’s now referred to as standard dynamic range + narrow color gamut + 8-bit depth) is minimal, in the general range of zero-20 percent when using the new HEVC video compression at direct-to-consumer bitrates… most definitely the Best Bang for the Bit. Compare this to 4K resolution: Even with the best video compression available today (HEVC), 4K resolution still requires somewhere around 250 percent the bandwidth of conventional HD (also coded using HEVC, for apples-apples comparison). And since the “bread and butter” of revenue still will be the conventional HD service, this bandwidth requirement most likely will be in addition to the existing service (that is, simulcast will be required).
IS 4K NEEDED? WHY NOT 1080P?
So in the case of bandwidth constraints, why not transmit in 1080p HDR+ and achieve the “wow” factor of HDR+? All 4K TVs upconvert 1080p to 2160p (4K) and the newer UHDTVs (4K + HDR+) will display an HDR+ image if the incoming stream has HDR+ coded, regardless of the spatial resolution. Many consumers may not even realize that the image was upconverted to 4K: An oft under-discussed issue is that the proper viewing distance for the human visual system to resolve 4K resolution is approximately 1.5x the picture height of a 16:9 display, whereas HD resolution can be resolved in full as far back as 3x the picture height. In most TV viewing environments today, consumers are sitting back from the display about 2.5-3.5x the picture height, therefore not really “seeing” much difference between HD and 4K.
Ericsson has done “unscientific” experiments at trade shows over the past two years, with two equal size flat screens side-by-side and we asked attendees which image they preferred (without telling them what they were looking at). Almost 100 percent would select the 1080p HDR display over the 2160p SDR display (and this was done at the proper viewing distance). This unscientific experiment is just one of many that have shown the “wow” factor of HDR+ to consumers, regardless of screen resolution.
So what about HFR? Like 4K resolution, HFR requires lots of changes in the studio and post production facilities. Its impact is also related to how fast or complex the scene motion is, so it’s extremely useful for high motion sports and nature documentaries, but doesn’t do anything much at all for “talking heads” and other low motion content. While there are examples of specialty content being shot at HFR—for video, this is defined as anything higher than 50 or 60 fps (country TV-standard specific), but it typically refers to 100 or 120 fps today—wider use of HFR likely will not occur for several more years to come, so it is not so big an issue for today.
In summary, if a provider is able to offer a UHD service that contains all of the five image technologies than definitely do it. This is just so much better than today’s conventional HD. However, if bandwidth constraints prevent “full UHD” from being delivered, the Best Bang for the Bit is definitely 1080p HDR+… with much of the “wow” factor and a far superior user experience over today’s conventional HD.