Q&A: Matrox Discusses Driving, Moving, and Streaming 4K

Matrox Graphics director of sales, Samuel Recine, shares his insight on the state of 4K video, and the challenges and opportunities surrounding encoding and decoding top-notch 4K streams.

CI Staff

When someone catches their first glimpse of an ultra-crisp, 4K display showing a full-resolution image, it’s a noticeably enthralling experience. The almost atomic level of detail capable with such high pixel count can create an authentic version of reality that hasn’t been realizable in the display market until now.

However, with any adolescent technology comes the question of how to create, share, stream, and receive a standard that has not yet achieved widespread adoption. For answers, Commercial Integrator approached Samuel Recine, director of sales (North America & Asia Pacific) at Matrox Graphics for his input as to what the infrastructure of a 4K world looks like and how we get there from where we are now.

What is currently driving 4K content in the market?

At the moment, most workflow 4K content is driven by PCs and cameras.

In PCs it is achieved in two ways: through rendering the content on a graphics card or decoding streamed 4K content on the PC’s hardware and software. Most modern graphics cards can deliver full 4Kp60, and power the full quality level and refresh rate of this pixel (display) density.

For the consumption of 4K in media and entertainment, game consoles and media players are increasingly shipping with 4K output resolution capabilities, and most models of commercial TVs can accept those inputs and serve up 4K viewing. Emerging smart TVs available on the market today support 4K both on the direct I/O side and on the stream decoding side.

How does moving SD or HD differ from 4K or 8K?

During the height of standard definition there was no real competition from IP versus traditional audio-visual infrastructures; it was prohibitively difficult/expensive to extend and switch video on IP.

Towards the middle of the transition to “High Definition” (or Full-HD), the market started to see a significant enough cost reduction in real-time encoding and uncompressed video to IP conversion for it to be viable to use IP as a medium to distribute quality video.

Real-time encoding became an “accessory to” AV infrastructure, supporting new features such as longer distances (over LAN, WAN, and the cloud), wireless devices (for example streaming HD to mobile devices), and hybrid environments (for example, nesting HD streams into Digital Signage Player PCs as part of the content being displayed).

In the 4K market (Ultra-UD, Digital Cinema, or 2160p), real-time encoding that can support full 4K (4:4:4 60Hz) at low bit rates, H.264 for example, are available right from the beginning of the shift to this new standard. And so are a brand new universe of products that can transmit, extend, switch, receive, mix, and display in the classic AV way of doing things.

For applications requiring absolutely no latency, such as live events, classic AV infrastructure and uncompressed-over-IP products still have the edge. But all other frontiers; from quality, to distance, to the low cost of switching, to wireless support and so on, are now also accessible with real-time encoding and AV-over-IP products.

If the products and technologies are available to integrators, why haven’t we made the shift to 4K seemingly overnight?

One issue I can see is that AV-over-IP AV companies are going to need to learn to differentiate between proprietary AV-over-IP solutions and non-proprietary AV-over-IP solutions. Proprietary often means a vendor is simply taking responsibility to provide an assured end-to-end experience by making all components: only their transmitters, their switches, and their receivers behave optimally together.

There is a compelling case for non-proprietary solutions given that 4K, low-bitrate, high-quality (4:4:4) full refresh rate (60Hz) real-time encoders already exist, and produce general-purpose streams with massive interoperability using the most widely deployed codec standard in the world, H.264.

They can encode for live streaming, AV-over-IP, and recording applications, as well as decode and display things like 4K cameras that stream their content. And all the required switching occurs on the same type of cables and switches as your networked computers.

As we first saw with SD to HD, and now HD to 4K, will these investments only be a temporary fix with the arrival of 8K starting a new technology cycle?

Understanding and shifting to quality compression over IP now makes sense because the underlying infrastructure (cables and switches) will be the exact same for the next node jump, which will be 8K. Today’s real-time 4K encoders can achieve as low as sub-50 millisecond (0.5% of 1 second) latency and 4K streams at full 60Hz can be transmitted in less than 20 Mb/sec.

When H.265 (also known as HEVC) matures and propagates to the same level of ubiquity as H.264 in the next 2 to 3 years, even 8K will fit in similar bitrates and latency levels.

Where does Matrox see itself in this universe of 4K content and AV-over-IP?

This year, we’ve been in this market for 40 years, and with a world-class team, our eyes are set on meeting the challenges that the surge of video brings to the business landscape.

Matrox is involved in both these topics through products that drive 4K content, and facilitate 4K AV-over-IP. From capture, encoding, streaming, recording, switching, managing, decoding, displaying, and video processing, Matrox products are capable of doing 4K at 60Hz at full 4:4:4 color depth in ultra-low latency and ultra-low bit rate.

Read Next: Virtual Reality is Coming to AV