On the last day of CEDIA 2017, Sony hosted an online-shashki Forum meetup in its demo room at the San Diego Convention Center. Sony Product Technology Manager Philip Jones presented abbreviated versions of two seminars offered by Sony on each day of CEDIA—this time, exclusively for online-shashki Forum members and their guests.
Both seminars included side-by-side comparisons of Sony displays (projectors in one, flat panels in the other) and comparable products from other manufacturers. Of course, when such a comparison is presented by one company, it must be taken with a big grain of salt. After all, who’s to say the presenting company didn’t stack the deck in some way to make its products look better? But as you’ll see, I don’t think that was a problem, at least with the flat panels.
The first presentation was all about projectors, which Sony claimed were in their best “out of box” picture modes with no other tweaking. Before we got to the side-by-side comparisons, we took a quick look at the VPL-VW885ES with an anamorphic lens. (The only other Sony projector that can accommodate an anamorphic lens in 4K is the VPL-VW5000ES.) The screen was a Stewart StudioTek 130 (204″ diagonal, 2.40:1, 1.3 gain).
The point here was to illustrate that using an anamorphic lens engages all 8 million pixels on the imagers, 2 million more than zooming the lens to “push” the letterbox bars off the screen, resulting in 30% more light on the screen. We also learned that the VW885ES can display 2.40:1, 1.85:1, and 16:9 4K images without moving the anamorphic lens, thanks to its powerful video processor.
Next, we looked at two projectors playing the same content on identical Screen Innovations Pure White 1.3 screens (160″ diagonal, 16:9, 1.3 gain). First up was the Optoma UHD60 (4K/UHD single-chip DLP, 3000 lumens, $2000) and the Sony VPL-VW285ES (true 4K, LCoS, 1500 lumens, $5000). Philip reminded us that the specified light output of projectors from different manufacturers cannot be compared directly, because each company uses its own measurement method, so the comparison is apples to oranges.
We started with a clip from La La Land on UHD Blu-ray from an Oppo UDP-203 player. The Optoma’s image looked noisy with flickering and a distinct color shift, while the Sony looked rock steady with virtually no noise and more natural color. Then we looked at a clip from Oblivion and focused on tone mapping. In the selected scene, Tom Cruise is tied to a chair in a dark room, and his shirt was clipped on the Optoma, but not on the Sony. Also, the spotlight in the background resolved to three lights on the Sony, while they blurred together on the Optoma.
When the Optoma was replaced with an Epson Home Cinema 5040UB (pixel-shifted “4K Enhanced,” LCD, 3000 lumens, $2700), the screen went blank. As Philip explained, that’s because the Epson’s HDMI ports operate at a maximum bitrate of 10.2 Gbps, and the player was sending 13.4 Gbps (24p, 12-bit, 4:4:4). I was amazed to learn that most 4K/UHD source devices, such as UHD Blu-ray players, Roku, and game consoles, actually send 12-bit 4:4:4 HDR, even though UHD Blu-rays, for example, store HDR at 10-bit 4:2:0. (This is the subject of an entire article on its own, which I’ll defer for now.) When Philip switched the player to 8-bit 4:2:2, the Epson was able to display it, but there was obvious banding in both displays.
Next up was a comparison of the JVC DLA-RS4500 (true 4K, LCoS, 3000 lumens, $35,000) and the Sony VPL-VW885ES (true 4K, LCoS, 2000 lumens, $25,000). Looking at a paused frame from Planet Earth II on UHD Blu-ray with the sun in full view, Philip measured the brightness of the sun at 126 lux from the Sony and 120 lux from the JVC. They were essentially the same, even though the JVC’s specified peak light output is 50% more than the Sony’s. This clearly demonstrated Philip’s point that the peak light-output spec from one company can’t be directly compared with the same spec from another company. In addition, the rock face in that shot had more detail on the Sony.
After that, we moved to an adjacent room, where Sony had set up several flat panels: Samsung QN65Q9F LED ($3500), Sony XBR-65Z9D LED ($4500), Sony BVM-X300 (30″ OLED reference monitor, $30,000), Sony XBR-65A1E OLED ($4000), and LG 65E7 OLED ($3500). All five were fully calibrated by Tyler Pruitt from SpectraCal, which put them on a level playing field. This is rare in a manufacturer-hosted side-by-side comparison, and I commend Sony for doing it. The sources were a Kaleidescape Strato movie server and a Sony UBP-X800 UHD Blu-ray player.
Philip pointed out that most HDR content is graded on an X300 (4K, 1000 nits) for color and detail and a Dolby reference monitor (1080p, 2000 or 4000 nits) for luminance. He also said that, for OLED TVs, the first goal should be to get as close as possible to what the image looks like on the X300—in other words, what the content creators saw.
Another point that Philip stressed was that each manufacturer implements its own tone-mapping algorithm; they’re not all the same. He also explained that the X300 and other reference monitors use no tone mapping at all. If the image clips on the X300, so be it; the luminance is graded on a higher-brightness display that does not clip. Sony believes that tone mapping on consumer TVs should not kick in until the brightness gets very close to the display’s peak luminance in order to maintain the director’s intent as much as possible. Others start tone mapping much earlier to reduce the potential for clipping, but this sacrifices brightness in other parts of the image.
To illustrate that point, we looked at a Sony test image with animated birds flying out of a top hat; the background was cyan, and the birds were a very slightly lighter shade of cyan. Looking at the two OLEDs, we could see the birds on the Sony but not on the LG, because of the difference in tone-mapping algorithms. You could turn down the contrast on the LG and see the birds, but that would screw up other things.
Finally, we looked at a clip from Wonder Woman. The first thing everyone noticed was that the black level on the LG was higher than the Sony A1E. Also, the color on the Samsung Q9 was not the same as on the other TVs. This was especially evident in an explosion.
Many thanks to Sony for hosting this online-shashki Forum meetup at CEDIA 2017, and especially to Philip Jones for his very informative—and fast-paced!—presentations. There’s so much to learn about HDR on projectors and flat-panel TVs, and Philip provided a lot of information to help educate us about the intricacies of this critical element in the brave new video world.