Video: Monolithic and Spine-Leaf Architectures

It’s hard to talk about SMPTE 2110 system design without hearing the term ‘spine and leaf’. It’s a fundamental decision that needs to be made early on in the project; how many switches will you use and how will they be interconnected? Deciding is not without accepting compromises, so what needs to be considered?

Chris Lapp from Diversified shares his experience in designing such systems. Monolithic design has a single switch at the centre of the network with everything connected directly to it. For redundancy, this is normally complemented by a separate, identical switch providing a second network. For networks which are likely to need to scale, monolithic designs can add a hurdle to expansion once they get full. Also, if there are many ‘low bandwidth’ devices, it may not be cost-effective to attach them. For instance, if your central switch has many 40Gbps ports, it’s a waste to use many to connect to 1Gbps devices such as audio endpoints.

The answer to these problems is spine and leaf. Chris explains that this is more resilient to failure and allows easy scaling whilst retaining a non-blocking network. These improvements come at a price, naturally. Firstly, it does cost more and secondly, there is. added complexity. In a large facility with endpoints spread out, spine and leaf may be the only sensible option. However, Chris explores a cheaper version of spine and leaf often called ‘hub and spoke’ or ‘hybrid’.

If you are interested in this topic, listen to last week’s video from Arista’s Gerard Philips which talked in more detail about network design covering the pros and cons of spine and leaf, control using IGMP and SDN, PTP design amongst other topics. Read more here.

Watch now!
Speakers

Chris Lapp Chris Lapp
Project Engineer, SME Routing
Diversified
Wes Simpson Wes Simpson
President, Telcom Product Consulting
Owner, LearnIPVideo.com

Video: Case Study FIS Ski World Championship

There’s a lot to learn when it comes to implementing video over IP, so it’s healthy to stand back from the details and see a working system in use to understand how the theory becomes reality. There’s been a clear change in the tone of conversation at the IP Showcase over the years as we’ve shifted from ‘trust us, this could work’ to ‘this is what it looks like!’ That’s not to say there’s not plenty to be done, but this talk about an uncompressed 2110 remote production workflow is great example of how the benefits of IP are being realised by broadcasters.
Robert Erickson is with Grass Valley specialising in sports such as the FIS Alpine World Ski Championships which were in the city of Åre in Sweden some 600km from Stockholm where Sweden’s public broadcaster SVT is based. With 80 cameras at the championships to be remotely controlled over an uncompressed network, this was no small project. Robert explains the two locations were linked by a backbone of two 100Gbps circuits.

The principle behind SVT’s project was to implement a system which could be redeployed, wouldn’t alter the viewers’ experience and would reduce staff and equipment on site. Interestingly the director wanted to be on-site meaning that the production was then split between much of the staff being in Stockholm, which of course was where most of the equipment was, and Åre. The cameras were natively IP, so no converters were needed in the field.

Centralisation was the name of the game, based in Stockholm, producing an end-to-end IP chain. Network switching was provided by Arista which aggregated the feeds of the cameras and brought them to Stockholm where the CCUs were located. Robert highlights the benefits of this approach which include the use of COTS switches, scalability and indifference as to the circuits in use. We then have a look inside the DirectIP connection which is a 10gig ‘pipe’ carrying 2022-6 camera and return feeds along with control and talkback, replicating the functionality of a SMPTE fibre in IP.

To finish up, Robert talks about the return visions, including multivewers, which were sent back to Åre. A Nimbra setup was used to take advantage of a lower-bandwidth circuit using JPEG 2000 to send the vision back. In addition, it carried the data to connect the vision mixer/switcher at Åre with the switch at Stockholm. This was the only point at which noticeable latency was introduced to the tune of around 4 frames.

Watch now!
Download the presentation
Speakers

Robert Erickson Robert Erickson
Strategic Account Manager Sports and Venues,
Grass Valley

Video: Efficient Carriage of Sub-Rasters With ST 2110-20

One of the main promises of IP video is flexibility and what better way to demonstrate that than stepping off the well-worn path of broadcast resolutions? 1920×1080 is much loved nowadays, but not everything needs to be put into an HD-sized frame. SMPTE ST-2110 allows video of all shapes and sizes, so let’s not be afraid to use the control given to us.

Paul Briscoe, talking on behalf of Evertz, takes the podium to explain the idea. Using logo insertion as an expample, he shows that if you want to put a small BUG/DOG/graphic on screen with a key, then real there’s not a lot of data that needs to be transferred. Typically a graphic needs a key and a fill. Whilst the key is typically luma-only, the fill needs to be full colour.

In the world of SDI, sending your key and fill around would need two whole HD signals and up to 6Gbps of data. When your graphic is only a small logo, these SDI signals are mainly redundant data. Using ST 2110-20, however, in the IP domain we can be much more efficient. 2110 allows resolutions up to 32,000 pixels square so we should be able to send just the information which is necessary.

Paul introduces the idea of a “pixel group” (pgroup) which is the minimal group of video data samples that make up an integer number of pixels and also align to an octet boundary. Along with defining a size, we also get to define an X,Y position. Paul explains how using pgroups helps, and hinders, sending video this way and then delves in to how timing would work. To finish off, Paul examines edge cases and talks about other examples such as stock tickers, not to mention the possibility of motion as we get to define the X, Y position.

Watch now!
This wall chart gives more info on pgroups and other low-level ST 2110-20 constructs.
Download the slides from this presentation

Speakers

Paul Briscoe Paul Briscoe
Principal Consultant,
Televisionary Consulting