8K was one important topic at the recent Samsung Advanced Display Summit. It's coming, and it offers several advantages over 4K beyond increased resolution.
Samsung and industry-analysis company Insight Media recently held a two-day technical summit about QLED and other advanced display technologies in Los Angeles, CA. The first set of presentations was all about 8K. I've long been a naysayer regarding 8K in the consumer marketplace, though my thinking on this has evolved lately, as I discuss here .
Four presenters—Chris Chinnock, President of Insight Media; Michael Cioni, SVP of Innovation at Panavision; Abhijeet Solat, Senior Engineer at Samsung Display; and Yoko Kimura, Sales and Marketing Manager at Astrodesign—offered compelling reasons to believe that 8K is coming and that it offers significant benefits over 4K.
Chris and Michael both began with the common objections to 8K. Aside from the notion that no one asked for it, Michael boiled it down to four main technical objections:
- The resolution is too high
- Bit rates are too slow
- Storage is too expensive
- Render times are too long
Michael addressed these objections in terms of historical progress. He framed his arguments using a unit of time he calls the "Moore cycle." As you may know, Gordon Moore, a co-founder of Intel, predicted that the number of transistors per square inch on an integrated circuit would double every two years or so. (Actually, Moore said 18 months, but it is often extended to two years.) Thus, in Michael's definition, a Moore cycle is two years.
He examined the historical progress of the four issues he cited in terms of Moore cycles. For example, from 2001 to 2009, video resolution went from standard definition (0.3 megapixels) to 2K (2.2 MP)—a factor of seven increase—over a span of eight years or four Moore cycles. Actually, the math is off by a factor of two; if resolution were to double every two years, video resolution in 2009 would have been 4.8 MP.
Michael continued the analysis to claim that video resolution reached 38.4 MP in 2015 (seven Moore cycles from 2001), which is closer to being accurate. This resolution is roughly equal to cinematic 8K (8192x4320 or 35.4 MP). We certainly saw 8K displays at the 2015 CES, though they were so-called "technology demos." Now, three years later, 8K TVs are coming to market, so we're a bit behind the theoretical curve.
According to Michael's hypothesis, we should be seeing prototype displays with 16K resolution by next year. In fact, he said he expects to see resolutions of 16K and even 32K coming to professional displays sooner than most might think.
Bit rates have increased more or less along a similar trajectory. In 2000, Firewire offered 400 Mbps, while in 2016 (and today), Thunderbolt 3 provides 40 Gbps. That's an increase of 100x over 16 years, not 256x as we would expect if Moore cycles are applicable to this parameter. Again, Michael's math is off by roughly a factor of two.
The cost of hard-drive storage follows Moore cycles more closely. According to Michael, hard-drive storage cost $18.40/GB in 2000, shrinking to $0.04/GB in 2018. That's nine Moore cycles, which should have seen a price drop of 512x. And his numbers reflect just that: $18.40/512 = $0.036.
Likewise, rendering times, which are dependent on computer-processing power, have increased dramatically. According to Michael, in 2000, standard-definition video was rendered in real time—that is, an hour of content took an hour to render. In 2018, SD rendering is 300x real time. Once again, the math is off by nearly a factor of two; according to his Moore-cycle hypothesis, SD rendering should be 512 times faster than it was in 2000. Of course, few people render SD any more, but it provides a level playing field when discussing the increase in rendering times over the years.
Michael's bottom line was that it doesn't cost any more to do 8K today than it did to do SD in 2000 (aside from some factors of two here and there). He and some of the other panelists also pointed out several benefits of 8K.
There will likely be very little native-8K content available to consumers for some time. (More on that in a moment.) Still, capturing images at 8K resolution results in better picture quality at lower resolutions.
Aside from the fact that downscaled images tend to look better than images captured at the final resolution, capturing at 8K provides greater flexibility in zooming and reframing at lower resolutions, which is especially important in sports. Also, a larger sensor lets the camera get closer to the subject with less distortion, and the depth of field can be even shallower than 4K allows, providing more creative possibilities.
Michael also claimed that capturing at 8K lowers the noise floor by 3 dB compared with 4K. And the noise floor decreases by 3 dB every time the number of "Ks" doubles, so 16K has 3 dB lower noise than 8K, and so on. This is a powerful argument for capturing images at ever higher resolutions.
Abhijeet Solat from Samsung Display pointed out that, aside from increasing the resolution by a factor of four, 8K increases the color gamut from DCI-P3 to BT.2020—or at least closer to BT.2020 than DCI-P3. This is made possible by quantum dots, which emit very narrow-band primary colors. He also claimed that 8K displays will reach a peak brightness of 4000 nits or more with 12-bit color depth, which supports 68 billion colors. In addition, 8K offers greater textural detail and perceptual depth.
One of the primary driving forces behind the push toward 8K comes from television manufacturers. As I mentioned earlier, we have seen 8K flat panels from various companies at CES for several years. At CES 2018, Samsung announced it would start selling the Q9S, an 85" 8K QLED TV, this year . And other TV makers are sure to follow in short order. On the projector front, Digital Projection demonstrated an 8K projector at NAB 2018 last April.
In fact, Abhijeet reported that Samsung Display—which manufactures raw panels for Samsung Electronics as well as other companies—predicts that 8K panels sales will reach 7 million units by 2020. This is much faster than industry-analysis firm IHS predicts, but he pointed out that 4K panel sales grew much faster than initially predicted as well.
At the other end of the production chain is the camera. Yoko Kimura from Astrodesign highlighted her company's AH-4801 8K camera head as well as Sharp's 8C-B60A 8K camcorder. Other 8K cameras include the Panavision Millennium DXL2, several Red models, Ikegami SHK-810, and Hitachi SK-UHD8000 series.
As we've seen at the NAB show for years, Japan's NHK public-broadcasting network has been developing 8K cameras and other elements of an 8K workflow. Likewise, Astrodesign now has a full line of 8K production equipment, including a 55" 8K LCD monitor and two 8K SSD recorders.
Connecting 8K devices is another issue. According to Chris Chinnock, single-cable connections capable of conveying 8K are now or will soon be available. U-SDI is a fiber-optic cable that will satisfy the needs of video and broadcast professionals. USB Type C, Thunderbolt, and DisplayPort 1.3/1.4 will satisfy the needs of the IT, professional, and commercial communities, and HDMI 2.1 will meet the needs of the consumer 8K TV market.
Content & Distribution
As you might expect, there won't be much native 8K content for a while, though experiments and trial runs have been conducted for several years, especially during live sporting events. These experiments have included all the Olympics since the 2012 Summer games in London, and this year's World Cup, which is being broadcast live via satellite in Brazil by Globo and Intelsat. NHK plans to capture some of the 2020 Tokyo Olympics in 8K and broadcast it via satellite in Japan, and Sony and Panasonic plan to have 8K TVs ready to sell by then.
It's unlikely that cable-TV services will implement 8K broadcasts any time soon—there is no 4K from these services yet, so 8K seems like a far-off dream for them. ATSC 3.0 over-the-air broadcasting can theoretically support 8K, but it would require significantly more compression to fit in a fixed bandwidth, so don't expect that in the near future.
Other than satellite, the other distribution avenue that can accommodate 8K "broadcasting" is online streaming. For example, YouTube added an 8K channel in 2015. But again, it requires severe compression to reach most homes, and few homes have an 8K display anyway. A better bet than real-time streaming is downloading 8K files, which you can watch after the download is complete. Hopefully, these files would use less compression than those intended for live streaming.
UHD Blu-ray does not currently support 8K, but Chris Chinnock reported that additions to the specification have been proposed or are close to approval. However, unless there is a significant increase in compression efficiency, a disc will be able to store only 25% as much content at 8K compared with 4K. So, I'm not holding my breath for that.
Make no mistake: 8K is coming, and sooner than most consumers expect. The push to bring 8K to the consumer marketplace is being driven mostly by TV manufacturers, who always want bigger numbers to use in their marketing. But camera makers and NHK are also barreling toward an 8K future.
The good news is that capturing and processing content in 8K is getting less expensive every year. Also, upscaling technology is getting better as well. This is critical, since most of the content delivered to 8K displays will be 4K and 2K, requiring high-performance upscaling. And as I learned at the Samsung summit, there are several important benefits to capturing in 8K, even if the end result is at a lower resolution.
Does this mean you should upgrade your new 4K display as soon as 8K models appear? No! The initial wave of 8K TVs will be exorbitantly expensive, so unless you're an obscenely wealthy early adopter, there's no need to hock the family jewels to get one. Just keep in mind that technology marches ever onward, and 8K will become more and more mainstream in the next five to 10 years. Once that happens, 8K TVs will be more affordable—and it might be time to upgrade your current display.