Graphics Reference
In-Depth Information
preliminary product implementations have already shown that HEVC encoding is
entirely feasible. Moreover, experience has also shown that as time moves forward,
the effectiveness of encoders to compress video within the constraints imposed by
the syntax of a particular standard can continue to increase more and more, while
maintaining compatibility with existing decoders. Indeed, encoders for MPEG-2
and AVC have continued to improve, despite the limitations of their syntax.
1.5
Profiles, Levels, Tiers, and Extensibility
Although we tend to think of a standard as a single recipe for guaranteed inter-
operability, some variation in capabilities is necessary to support a broad range of
applications. In HEVC, as with some prior designs, this variation is handled by
specifying multiple “profiles” and “levels”. Moreover, for HEVC a new concept of
“tiers” has been introduced. However, the diversity of separate potential “islands” of
interoperability in version 1 of HEVC is quite modest—and depends on the intended
applications in a straightforward manner. Only three profiles are found in the first
version of the standard:
Main profile
: for use in the typical applications that are familiar to most
consumers today. This profile represents video data with 8 bits per sample and
the typical representation with a “luma” brightness signal and two “chroma”
channels that have half the luma resolution both horizontally and vertically.
Main Still Picture profile
: for use as still photography for cameras, or for
extraction of snapshots from video sequences. This profile is a subset of the
capabilities of the Main profile.
Main 10 profile
: supporting up to 10 bits per sample of decoded picture
precision. This profile provides increased bit depth for increased brightness
dynamic range, extended color-gamut content, or simply higher fidelity color
representations to avoid contouring artifacts and reduce rounding error. This
profile is a superset of the capabilities of the Main profile.
However, the syntax design is highly extensible, and various other profiles
are planned to be added in future extensions. The extensions under development
include major efforts on extensions of the range of supported video formats
(including higher bit depths and higher-fidelity chroma formats such as the use of
full-resolution chroma), layered coding scalability, and 3D multiview video. The
JCT-VC, and a new similar organization called the JCT-3V for 3D video work, have
continued to meet at the same meeting frequency to develop these extensions and
they remain very active in that effort—with more than 150 participants and more
than 500 contribution documents per meeting.
While profiles define the syntax and coding features that can be used for the
video content, a significant other consideration is the degree of capability within a
given feature set. This is the purpose of “levels”. Levels of capability are defined
