High Dynamic Range broadcasting
Watching TV is a natural part of our lives. Since the first TV sets made their way into our living rooms several decades ago, the TV experience has gone through several steps of evolution: from black-and-white to color TV, from analog to digital, and recently an increase in resolution from standard definition to high definition and then even further to 4K (a.k.a. Ultra-HD).
The next big thing is High Dynamic Range, HDR.
HDR not only brings the opportunity to better represent bright parts of a scene but also provides an improved level of detail in the dark parts. The end result is a more immersive viewing experience.
Please have a look at this video we made, where we explain HDR and some of the challenges associated with it.
The video group at Ericsson Research – along with Ericsson’s TV & Media R&D team and technologists – have been working actively to figure out how existing TV infrastructure can be used for HDR content. In TV broadcast production, video sequences are captured by different cameras in the studio and in the field and will go through several different steps of processing before reaching the end user. The processing steps are applied in order to provide good video quality while still meeting the constraints of the different interfaces in the delivery chain. Ideally, you would like to reuse as much as possible of your existing processing elements but perhaps configure them differently in order to reach optimal results for HDR.
The last step of this process is how to convert the digital bits into light, this is known as the electro-optical transfer function (EOTF). Some new TVs on the market supports “Ultra HD Blu-Ray® video characteristics”, which means they have an HDR-friendly EOTF. The Ericsson approach is to deliver content in this format, so that no part of the receiver needs to be changed in order to receive HDR. In order to make that look good however, some clever tricks must be applied before encoding of the video, and we at Ericsson Research have been developing methods for this. The image below shows the difference between the conventional processing and the processing we are promoting.
Detail of an image from the Market sequence, courtesy of Technicolor and NEVEx (Next Video Experience).
To the left we see the original image. In the middle we see what happens if we only change the transfer function, without the special processing – artifacts appear. To the right we see what happens if we change the transfer function and apply Ericsson’s special processing.
Looking ahead, our focus is to ensure that descriptions of how to apply efficient HDR processing are publicly available and that standards and specifications for HDR do not introduce unnecessary complications for existing implementations. To accomplish this, we are active in a large number of standardization bodies: MPEG (ISO/IEC), VCEG (ITU-T), ITU-R, SMPTE, DVB, ATSC, etc. These groups are all looking into HDR from different perspectives.
One key question is whether there will be several different solutions for HDR in the market or if a single approach can encompass all demands. We’ll make sure to keep you posted on where things are going!