What business and/or technical challenges is JPEG XS designed to solve?

By Simen K. Frostad

Chairman of Bridge Technologies

Q&A published by CSI Magazine

What business and/or technical challenges is JPEG XS designed to solve?

The technical challenge for JPEG XS is one that constitutes an age-old and ongoing challenge for the industry as a whole; how to transport super-high-quality data from venue to venue efficiently, using as little bandwidth as possible, and ensuring the highest speed, the lowest cost, and the greatest visual fidelity.

JPEG XS represents a remarkable solution to this issue. It constitutes an interoperable, visually lossless, low-latency and lightweight image and video coding system, with a compression ratio typically around 10:1 depending on the resolution and content type (ratios vary from about 7:1 – 14:1).

How would you assess the takeup so far, in terms of the ecosystem and deployments?

The take-up for JPEG XS has been extensive – indeed, I’d describe it as phenomenal: few standards receive such universal acclaim and quick adoption. But this really speaks to the technological strength of the JPEG XS standard; it represents the pinnacle of transporting high quality video. The triangular relationship between latency, quality and bitrate would normally mean that as production quality starts to increase into the realms of 4K and 8K, there is a corresponding impact on delay and bandwidth requirements. JPEG XS represents a remarkable method for circumventing this; allowing for exceptionally high quality to be delivered in real time without a corresponding need to increase the infrastructure of the network – thus avoiding the tremendous costs that would be associated with that. It largely eliminates the trade-offs that were once inherent when it came to the relationship between size, quality and speed.

The standard – combined with ST-2110-22 – facilitates huge changes in the way that remote production is undertaken, not just in terms of OB vans, but the potential to move entire fixed operations to data centres. And the vendor interoperable nature of the standard means that it’s open to everyone, and everyone producing technologies for the industry can leverage this tool within their own products to facilitate a joined-up environment for broadcasters that maximises the potential they can reach in terms of live production and broadcast at the highest levels of quality, with a reduced OpEx. 

How is your company involved in this space?

Analysis of transport streams encoded with the JPEG XS standard is facilitated by both our VB330 and VB440 probes, meaning that we are able to provide operational insight in both production and distribution spaces.

With the VB330, all of the existing functionality that the probe provided for other formats has now been extended to the JPEG XS standard, across both appliance and software versions. This means that up to 2000 multicasts using the JPEG XS standard or up to 50Gb of capacity can be monitored, with comprehensive deep-dive metrics of the data presented in an intuitive, usable manner; including measurements of bit rate, packet drop and excess jitter, as well as continuous thumbnail decode for validation of the data. Alarming across all streams and measurement metrics can also be set.

More than this though, the Timeline ‘content’ option has also been extended to cover JPEG XS streams. This allows for both thumbnails and measurement metadata to be recorded directly to the appliance for up to four days (or in the case of the software probe, extended through disk expansion). This means that data can be held and archived locally – including visualisation of thumbnails and identification of Quality of Experience issues such as blackframes, freezeframes, sound intensity and MOS scoring.

In relation to production, the fact that the VB440 operates in accordance with ST-2110-22 means that it naturally incorporates the ability to monitor data compressed with the JPEG XS standard. As such, all of the extensive capabilities delivered by the VB440 can be used by creatives and directors alike to facilitate in-the-moment insight, regardless of whether the data is compressed or uncompressed. 

Overall, what can we expect in the future over the next few years (e.g., tech evolution and takeup)?

The balance between expressing industry insight/foresight and the hubris of prediction is a difficult line to straddle, and so I’m cautious to make any bold claims. JPEG XS is doing everything it should be doing, and the industry is responding to it positively. It’s so extensively used that there is not really much more to occur in terms of take-up – except if you count those who have not yet made a move to IP at all, and when they do, the adoption of JPEG XS within that space will be all-but-guaranteed. Other standards are working hard to maintain their presence in the space and evolve their capacity – some targeting niche sectors or application, but in the field of broadcast, it will be JPEG XS that maintains dominance.