A simple answer to the above question might be; just increase both spatial and temporal resolution. It is not that simple.

Next generation TV sets will be able to display 4K images at higher frame rates with improved contrast ratios and larger colour spaces. These benefits will be marketed by set manufacturers to keep their factories operating and profits up. Content owners hope to charge a premium for brighter and sharper productions. And finally, consumers want larger displays to showcase this newer technology.

So it is going to happen, but what are the difficulties and what challenges need to be overcome to make UHDTV as ubiquitous as HDTV today?

Challenges in the production pipeline

The 4K and higher production issues will resemble those that had to be addressed with the transition to HD.

Light: As the number of pixels goes up, unless you increase the sensitivity of the sensing elements, the amount of light on the subject has to be increased. Camera sensors are unlikely to suddenly be four times larger.

What is happening is that the S/N ratio of the image capturing system is improving.

However, as Figure 1 shows, increases in sensor sensitivity have not kept up with decreasing pixel sizes.

Focus: Camera operators often use autofocus. Sometimes it works well, but you can really see the result when it fails in 4K. Professional 4K camera manufacturers offer different types of “Focus Assist” solutions, but it remains to be seen how useful these will be in the heat of the shoot.

Set Design: Set details that may not have been visible in HD will now be apparent. This is even more evident with the transition to HDR. An object that had been lost in the shadows will now be visible and a seam on a white wall that might have been invisible will now ruin the illusion.

Lenses, as John Watkinson has made abundantly clear in his series Camera lenses: Through the looking glass, Part I, are most likely to be the limiting factor for sharpness and a large contributor to the sensitivity in 4K imaging. The result of the compromises being made to fit 2/3” lenses to 4K imaging chips, taking advantage of the sharper MTF falloff, result in a 1/4 screen HD cutout with lower quality than that available from a dedicated HD camera. While this may be an acceptable compromise, everyone should realize the tradeoff being made.

The camera “data delta” we spoke about in Tips to Selecting a Video Camera, Part 1 will present many challenges going forward. Remember that using intraframe codecs (ProRes, DNxHD, etc.) was only necessary to make editing simpler. With increased compute power this is no longer necessary and interframe codecs certainly give a better picture at the same data rate.

The question of how the data stream coming off the sensor is processed for presentation at the viewfinder will take some deep thought. We do not want the picture to look different than that presented by the post processing systems. Perhaps an optical path to the viewfinder will be the lesser of two evils, i.e. we accept the reduction in sensitivity due to light loss in exchange for a more accurate representation of the scene being shot. Another option is to have an ultrasharp B/W viewfinder and trust that the camera has been setup correctly. Will camera operators accept such a solution?

At the end of the day it may not be possible (at least for the next few years) to get true 4K HDR HFR performance from a shoulder-mount sized camera. This may not matter for a system camera mounted on a large pedestal or a wired camera head on a track or cable traverse above the stadium. As Mark Grinyer at Sony has said “ It's obviously very early days for 4K in live production and all this detail needs to be learned, as it had to be for 3D."

Part 1 of my series, Getting from HD to the UHD Perfect Picture, I spoke about the data delta that is inside every camera and the challenges that represents. Remember that sticking to intraframe codecs (ProRes, DNxHD, etc.) was only necessary to make editing simpler. With increased compute power this is no longer necessary and interframe codecs certainly give a better picture at the same data rate.