High Dynamic Range (HDR) was the hot topic at IBC in Amsterdam this year. But as visually exciting as HDR can be, there are a lot of checkpoints along the way that block progress and could interfere with the speed of adoption.
HDR refers to the extended range of light possible in the image. It requires that a larger range of luminance values be transmitted or stored, and that a larger range of luminance be displayed than found in standard range displays. A new term, HDR+, refers to the combination of HDR with wider color gamut. The DCI P3 color gamut associated with digital cinema is already larger than that available for HDTV. But larger gamuts are now possible, and best associated with higher levels of luminance, thus HDR+.
With a larger range of luminance to capture, one would expect HDR to require more bits to represent the signal. One of the areas where competition exists is in how to best encode the wider luminance range so as not to require more bits. These functions are called the Optical-Electro Transfer Function (OETF) and Electo-Optical Transfer Function (EOTF). Proposals from Dolby, Philips, and BBC offer incompatible OETF/OETFs, some offering a degree of backwards compatibility from HDR to SDR.
To add to the soup, a variety of transmission and storage methods have emerged: Single delivery, scalable methods such as Dolby’s dual-layer scheme and MPEG SHVC (scalable HEVC), and the single-layer backwards compatible stream proposals from BBC, Philips, and Technicolor.
And you thought immersive audio was messy.
The strategy behind the single-layer formats is simple. A single disc, download, or broadcast is designed to produce acceptable images on all displays, whether standard dynamic range (SDR) or HDR. Backwards compatibility through a single layer generally requires signal plus metadata, where the metadata is read by the more capable display to produce the HDR (or HDR+) image. This requires the metadata to do a lot of work, as both luminance ranges and color gamuts must vary depending on the capability of the display. Given that more than one backwards compatible format has been proposed, it can be assumed that some formats work better than others. Technicolor is the player in this group that appears to be attracting significant attention.
Dual layer formats are designed to reduce or eliminate compromise, relying less on metadata and more on signal. The first layer carries the SDR image, while the second layer carries the additional information needed for high quality HDR+. For content owners, the dual layer approach has the advantage to enable high quality HDR for premium-priced Blu-ray discs. Dolby Vision is the leader in this pack.
The diversity of proposals for consumer HDR is largely driven by the need to support backwards compatibility. However, it is unlikely that backwards compatibility will be a driver in the HDR version of digital cinema. Dolby Vision for cinema uses the 12-bit image data format, but with a different EOTF. While there are a few EOTFs that could weave their way into cinema, the professional industry appears to be circling in on the Dolby PQ EOTF, which is now standardized as SMPTE ST2084.
The complexity of consumer HDR will very likely slow down its adoption. Consumers require simplicity, and right now HDR appears to be anything but. This would be good news for cinema, where a consumer appetite for HDR images could be quite difficult to satisfy. After all, the only HDR cinema projector that currently exists is the Dolby Vision projector, available only in Dolby Cinemas. But this can also work both ways, and Dolby may be counting on it. If the only experience of HDR in cinema that consumers experience is from Dolby, then perhaps that will help Dolby Vision shine through the clutter on the consumer side. If so, it would be an impressive revival of the top-down strategy of using cinema to drive technology to the home.