Editor's Eye

If you ever wonder what the geeks at Sound & Vision do when we’re not listening to new speakers or tuning up video displays, well, we’re probably debating some arcane technical detail that most non-enthusiast mortals would neither understand nor care about. And so it was that a rather fired-up exchange of e-mails occurred recently between myself, video technical editor Tom Norton, and our contributing technical editor Kris Deering.

What sparked the discussion was Kris’s mention in a review manuscript that a particular projector under test touts support for the P3 color gamut, but that this is not useful to our audience because there is no P3 consumer content — he pointed out that 1080p HD is mastered in Rec. 709, while 4K/UHD content is mastered in Rec. 2020. When we sent the manuscript to Tom for his technical fact-check duties, he flagged this for query, acknowledging that, technically speaking, UHD uses a Rec. 2020 “container,” but that no consumer displays can currently show full Rec. 2020 color and that UHD Blu-rays use the more limited P3 gamut within the Rec. 2020 container. No, Kris replied vehemently, P3 has nothing at all to do with consumer video, and there is no place for even the mention of it in a discussion about today’s consumer UHD content and displays.

But then, why is P3 so frequently cited in the marketing and reviewing of both UHDTVs and the UHD Blu-ray format?

I saw from the get-go that their dialogue was less about any disagreement on the technical details and more about semantics. And our discussion, including a couple of clarifications from Imaging Science Foundation’s Joel Silver, brought me to the admission that we’ve been guilty of regularly practicing some misleading language when we discuss color gamut.

For those not fully versed, we use that term to describe the maximum range of colors that a video signal can carry or that a display can visibly produce for your eyes. This is not to be confused with a display’s color bit depth: typically 8-bit in 1080p displays, more frequently 10-bit in newer HDR-capable UHD displays, and occasionally 12-bit with some high-end displays. Bit depth affects the number of gradations of color that can be achieved within a given gamut; more bits means finer gradation and less potential for visible banding artifacts. But it does not affect how deeply red, green, or blue the display’s primaries can be pushed when called upon to do so.

Those limits are defined loosely (but not fully) by the two dimensional coordinates on a CIE color chart like those shown in the diagram above (courtesy of ISF). The humpback CIE chart represents all visible colors, and the highlighted triangles represent each of the gamuts. You’ll notice that DCI-P3, the gamut used for mastering movies distributed to digital theaters, extends much deeper into the saturated green and red areas than the Rec. 709 gamut we’ve used for years in HDTV. It’s a quite visible difference that allows red, in particular, to look much more like red and less like the orange-red we’ve seen prior to now. Rec. 2020, meanwhile, extends the boundaries even further.

Today’s consumer UHD displays offering wide color gamut as a feature are often defined by the percentage of P3 they can achieve — 90 or 96 percent of P3, and so on. Similarly, we speak of UHD Blu-rays as being mastered to the P3 standard. But the reality is that UHD consumer displays wouldn’t look right showing a signal mastered in P3, nor is any UHD content actually mastered in P3. These are all Rec. 2020 displays and Rec. 2020 content.

The constant references to P3 in describing UHD Blu-rays or other UHD movie content stem from the fact that most begin with a DCP (digital cinema package) that’s genuinely mastered to P3, using a professional P3 monitor. This must be modified in some key ways and have its RGB values recoded to be placed on a Rec. 2020 UHD Blu-ray or streamed to a UHD display. What you end up with is content mastered to Rec. 2020, but which fails to fill the full Rec. 2020 container beyond where the P3 primary color points would typically lie...if it were mastered to P3.

For now, this is the widest color the studios will master to because (a) it makes life easy for them given that they’re already preparing a P3 DCP for cinema distribution, and (b) our consumer displays can’t yet reproduce colors much beyond the P3 limits anyway. For their part, the TV manufacturers ought to be telling us how close their sets come to reproducing full Rec. 2020 gamut since these are Rec. 2020 displays. But in the numbers game that is TV marketing, a TV maker would much rather tout that its UHDTV can hit 96 percent of P3 than have to say it’ll only do 70 percent of Rec. 2020 — especially when there’s no content out there yet that pushes out to the Rec. 2020 limits. Technically, though, only a characterization of Rec. 2020 coverage is truly accurate for a consumer television. You can read more about this in Tom’s A/V Veteran blog Colors in Space.

In a way, both Kris and Tom were right, and both understand the technicalities at play. Tom recognized that UHD content is mastered to Rec. 2020 but wasn’t really wrong in characterizing the gamut on a UHD Blu-ray as P3, as that reasonably describes the limits of its color implementation and loosely beckons to its origins in digital cinema. Nor was Kris wrong in insisting on precision in our description, though it may be too rigid to suggest that P3 has no place at all in discussion of our current UHD content and displays. Suffice to say that the devil, as usual, is in the details. We’ll try to be clearer going forward when discussing these concepts in our reviews.