Cost efficient, lossless transport of uncompressed video using RIST
By Alexander Sandström, Head Of Product Marketing - Streaming Solutions, Net Insight
One of the key benefits of RIST is that it delivers lossless broadcast-grade quality over any type of infrastructure. Historically, only dedicated fiber networks and satellite links have provided good enough quality for professional broadcast workflows such as contribution, remote production and primary distribution. For most organizations, the main reason for adopting RIST has been as a complement to satellite or fiber, to provide lossless media transport over the public Internet. And in practice, while RIST makes internet transport lossless, the bandwidth limitations of most internet connections in the past means that this type of solution has mostly been used for highly compressed content.
But Internet bandwidths are increasing fast, with 100mbps and Gigabit Internet connections becoming mainstream. On top of that many broadcasters and professional media organizations are starting to use standard IT/enterprise leased lines as a replacement for dedicated fiber to reduce cost. Pair RIST’s lossless transport with cheap high bandwidth links, and you open up for lightly compressed or even uncompressed content to be transported extremely cost efficiently.
Limitations yesterday, solutions today
Visually lossless compression schemes like JPEG2000 use something like 150/250/1000Mbps for one HD/3G/UHD feed, mathematically lossless compression like JPEG-XS may use something like 350/700/3000Mbps for one HD/3G/UHD feed and of course uncompressed uses 1500/3000/12000Mbps. In the first incarnation of RIST, the Simple Profile (TR-06-1), standard GRE tunnels were used as the foundation, and they came with limitations in how much data could be kept in flight between sender and receiver at the same time. The amount of data kept in flight is a product of the bandwidth transported and the latency between sender in the receiver. Meaning that with the simple profile, lightly compressed or uncompressed content could only be transported over network links with very low latency.
But this limitation is just about to be solved with the upcoming RIST Main Profile (TR-06-2). The Main Profile continues to use standard GRE tunnels, but uses extended sequence numbers to increase the amount of data that can be kept in flight between sender and receiver. What this means is that with RIST Main Profile, lightly compressed and even uncompressed content can be transported over any type of network link no matter its latency. And the good thing is that this goes also for upcoming video formats using even more bandwidth, such as 8K and even higher resolution 360/VR formats, making RIST Main Profile incredibly future proof.
Table: Examples of sustainable round-trip time (RTT), i.e. two-way latency, for different lightly compressed and uncompressed video quality levels.
Light compression in cloud workflows?
What makes things perhaps even more exciting is that now that we can transport lightly compressed and uncompressed video over cost-efficient public links, we can also start pushing such content into cloud environments for processing. In the past, cloud-based workflows have used H264-based mezzanine formats, with bandwidths in the range of 50Mbps or so. In the future, with RIST Main Profile, you will definitely see lightly compressed JPEG 2000 mezzanine formats being used in cloud workflows. And don’t be surprised if, further down the road, you see even lighter compression being used in the cloud, thanks to RIST Main Profile.