Planning for the 4K transition

Tuesday, March 4, 2014
When we look back to the dawn of consumer video, we always had to choose between different formats and brands. How many people today remember the VHS versus Betamax home video market war?
Professional broadcasters face the same challenges. The standard definition (SD) to high definition (HD) transition is still in progress. Delivery formats change, distribution of content is more versatile than ever: terrestrial, cable, satellite, IPTV and OTT and so on. 
In recent years we have witnessed the introduction of stereoscopic television. It didn’t get the expected market share and its popularity sharply dropped after the first impression of cool novelty. The lack of meaningful content and the constrains of the viewing environment made it something to try but not to experience on a daily basis.
The trend now is towards 4K/UltraHD television. Is it just the next buzz word or is it here to stay?  4K/UltraHD is roughly four times the 1920 x 1080 pixel resolution of HD and delivers much better picture quality. It brings new challenges to the broadcast workflow: four times the
storage capacity, four times the processing power, four times the delivery bandwidth, four times the rendering time. 
The best thing about 4K/UHD is not just the resolution but the parameters defined in the new ITU.R –Rec. 2020: larger colour space and progressive-only frame rate. This delivers the much needed update of the current Rec. 709 (first approved in the early 1990s). Dealing with progressive-only frames finally cuts the holdover from the analog broadcast use of interlaced video. Rec.2020 specifies - Picture temporal characteristics: Frame frequency (Hz) 120, 60, 60/1.001, 50, 30, 30/1.001, 25, 24, 24/1.001. Scan mode: Progressive.
ITU.R – Rec. 2020 identifies D65 as the white point of its colour space, located at [0.3127, 0.3290], the same as in Rec. 709. The differences are the available gamut and the fact that Rec. 709 is limited to 8-bit depth, whereas Rec. 2020 supports 10- and 12-bit depth. This apparently equates to a 75.8% coverage of the CIE 1931 colour space compared to Rec. 709′s 35.9% coverage.
All of this indicates that the 4K/UHD is a step in the right direction and is here to stay. The question is: How soon we need to be ready to support 4K/UHD in our workflow? The answer is different for the various stages of production, post-production, playout and delivery. 
In the production case, adopting the technology immediately makes the most sense. It is a cardinal rule always to make acquisition of content at the highest possible resolution, frame rate etc. Various manufacturers offer 4K/UHD equipment which is now widely used in modern digital cameras. Down-converted 4K is used for monitoring during shooting due to the lack of proper 4K monitoring equipment.
Post-production is a little different. Storage, CPU and rendering time requirements are four times higher; HD proxies are normally used during editing. Original material is always available to conform the final edit when 4K/UHD output is necessary. The higher resolution of the source material is very helpful when compositing or if other forms of content manipulation are needed.
Playout is one of the most challenging stages due to the variety of requirements and the amounts of metadata needed to co-exist with the media assets. Starting from the ingest part, this is a completely different task from the camera content acquisition. Typically we ingest from satellite feeds or files and it is essential to preserve all metadata embedded and accompanying the content.  Ancillary data such as AFD, closed captioning, time code is also very important because changing of formats, aspect ratio and up/down conversions need to be performed during playout. Audio language descriptors and loudness/true-peak information are equally important in order to assign the audio languages to the outputs properly and to obey the mandatory loudness requirements.
The most important characteristic of a playout system is that it should be built on a format agnostic engine. It must also be easy to add new input or output plugins to handle the desired format while the core logic remains the same. This allows for unchanged workflow from a user prospective.  
Such an engine can also scale character generation and graphics to the needed resolution, allowing the mixing of all available content in a single playlist. Re-purposing of content is one of the most common requirements in television broadcasting and is made possible by rescaling/frame-rate-converting HD content which can be used together with native 4K/UHD content. Real-time processing is essential.
Monitoring is also very important. The 4K monitoring solutions offered on the market don’t always cover the whole workflow which needs to include distribution of the signal to multi-viewers for example. A flexible playout engine can deliver different formats in parallel and accommodate the current infrastructure for monitoring while delivering the 4K signal.
Delivering the content to the end user is another challenge and one which is currently the biggest obstacle 4k/UHD broadcast meets. The bandwidth requirements are hard to overcome in the technical and economic sense so new compression codecs like the HEVC-h.265 are expected to evolve and contribute to the adoption of 4K.
Big worldwide TV broadcast events like the 2014 Winter Olympics in Sochi and the 2016 Olympic Games in Rio are scheduled to be broadcast in 4K. The Tokyo 2020 Olympic Games are planned to be broadcast in 8K. 
In conclusion, the most important part of planning any upgrade is to future-proof the infrastructure. During the transition period, mixed format content and metadata will need to be used and co-exist in the same environment. Format handling and signal flow are one aspect but equally important are the workflow changes from the user prospective. Adding extra steps always lead to issues with productivity and the necessity for additional training which may be significant obstacles to overcome.