In Pursuit Of A Better Live Production Switcher
Production switchers are now being tasked with handling an ever-increasing number of sources that can stress switcher resources.
Like many critical professional equipment categories, live production switchers (“vision mixers” to many overseas) have undergone a transformation based largely around the implementation of software-defined features and multi-layered capabilities that allow the operator to do more. When EVS introduced its Dyvi switcher in 2017, it boasted a software-defined architecture - basically a software approach to utilizing GPU performance to manipulate and modify video streams - that offered users a brand-new approach to live production while promising to take them far beyond the limits of conventional hardware switcher design.
EVS VP Product Manager John Carter sat down with The Broadcast Bridge to discuss how the switcher operator - or technical director (TD) - community has evolved, the new challenges they face to support increasingly more devices, and how the Dyvi has been engineered to meet those challenges.
The Broadcast Bridge: What are today’s TDs looking for in a video production switcher? What features matter most?
John Carter: With the different types of production genres and TDs individual creative approaches to the variety of productions, the core features that are relied on still remain the same in the general context. For example: the way camera shots are called, images mixed and transitioned, complex composites augmented with graphical elements and the guaranteeing of what is allowed to contribute to a single or multiple outputs.
What is changing is new entertainment types, the format of these live productions and, of course, production costs. A typical example here is producing eSports. These newer shows and their multiple formats can use a huge amount of switcher resources, such as the keyers, DVE and internal clipstore capacity. Instantly recalling effects that support unscripted action or reaction can quickly use up all the available resources within the traditional production switcher.
So while these general features are still required, we are seeing the demand and the quantity increase. TDs need a solution that meets a particular show’s demands with an adequate amount of outputs, but they also need to create a workflow to their individual style of running the show regardless of how complex.
The Broadcast Bridge: The Dyvi is basically all software being rendered on a GPU card. Explain why that is a benefit.
Carter: One of the biggest benefits achieved with this approach is you make all the required composites and effects simply and work anyway you want. The GPU element is presented with the required sources and what you desire on the outputs. You don’t tie up your desk or resources and this approach gives you the license to go anywhere you want to with your next transition; stress free.
Also at the key layer you’re not constrained to a prescribed set of functionality, as the ability to manipulate every layer is the same with the Dyvi. As an example, you’re not contained to a number of keyers that can only be resized. Also, you don’t have to think cascading and therefore avoid quickly tying up needed resources that are already on air.
The Broadcast Bridge: The Dyvi allows users to set up and launch effects in a whole new way. Describe the process.
Carter: As part of any show build or preproduction time, effects must be built, stored and recalled easily. With a layering approach to building effects, the Dyvi GUI is designed to encourage more of a compositing mindset. TDs use a mouse to navigate around the GUI menu structure on the control surface. It offers a straightforward contemporary method to building and grouping layers together. You don’t need to think of signal flows through the desk for a particular effect, or for a particular point in the show, and build around this. More importantly, you will have the resources free and available to you during show time when you need to call the supercool move.
In term of launching effects, the Dyvi control surface is laid out in a traditional way to allow the confident M/E operation for the show switching tasks, however the panel is totally customizable and can be treated like a shot box. You can call up at will what you’ve built. Any of your effects can also be assigned as a source on the busses and be switched through the transition area. There are no restrictions as to where or how the TD wishes to switch.
If you’re in the live automation environment, you simply create the effects for the show design brief and the switcher’s Automation system will call these up quickly and transparently.
The Broadcast Bridge: The switcher uses Processing Modules (PMs) and proprietary I/O cards and a 2RU COTS server that operators use to build different configurations for different projects. (For example, five PMs can be stacked to create a 192x96 HD switcher.) Why is this “virtualized M/E” approach better than other switchers that use FPGAs and ASICS to create pathways for rendering or processing video?
Carter: The Dyvi is a scalable solution, which allows single or multiple PMs to support multiple configurations. It can be seen as a fluid “Production Core”, with the size of this video-processing core defined by the numbers of PM that you desire to have in it.
For a single large one-off show - where there are a lot of inputs, outputs and simultaneous effects - several PMs can be grouped together. For smaller shows the production core can be simultaneously structured accordingly. As this is a software-defined approach, and in combination with IP infrastructures, the Dyvi can easily be configured to be a substantial production core or smaller separated production cores with great creativity capabilities.
With proprietary hardware designs implemented in FPGA technology, there is a ceiling that is hit, so you can only go so far. This limit is determined fundamentally by the backplane capacity of the mainframe supporting the number of busses, M/Es, etc. The Dyvi approach using COTS equipment allows interfacing to multiple PMs and users can easily increase headroom if required.
The Broadcast Bridge: Because every Layer can be manipulated does Dyvi run in a constant DVE mode, and does introduce variable delays?
Carter: With the GPU approach, it’s not so much a DVE mode but the intelligent handling of sources and outputs that are required at any specific time. The important point here is that everything is always synchronized regardless the size of the Production Core. Any source, regardless of what processing is placed upon it and how many times it is being used, will always be in time with itself and other sources, even if it is in the same effect simultaneously.
Part of this benefit is when working in multiple PM systems everything is timed together. Even if other traditional approaches could transfer sources between their mainframes, the delay would increase and it would not be consistent depending on what signal paths are switched up. The Dyvi is currently used for many live projects, with many customers proving that this approach works and has been accepted.
The Dyvi today is UHD capable and as part of the wider roadmap will shortly be processing native UHD full raster.
The Broadcast Bridge: It’s been said that on a traditional switcher you run out of M/Es and keyers, while on the Dyvi (which is a GPU-based switcher) you run out of time, which can cause sync issues. The Dyvi actually includes a meter to show how much you are out of sync. How do operators get around this?
Carter: With traditional approaches, you will either hit the limit as discussed earlier or you will be tied up with the resources that are in use, constraining the workflow through the desk. The Dyvi does indeed provide a confidence meter to what level of GPU processing is being used, however if there is a condition that raises concern to the amount of GPU processing, then any concerns can be off loaded to be processed by another PM in the Production Core and brought in as a feed, which is all transparent to the TD at work.
The Broadcast Bridge: Can the Dyvi handle 4K UHD (12G) and HD (3G) productions simultaneously? How?
Carter: The Dyvi today is UHD capable and as part of the wider roadmap will shortly be capable of processing native UHD full raster. As the fundamental Production Core can process progressive video, scaling down from UHD to HD is supported. It is an evolving product approach and will take advantage of COTS solutions and the benefits that these technologies bring for formats of the future.
The Broadcast Bridge: Many sports broadcasters are now doing remote “At Home” productions. How does the Dyvi complement this type of workflow?
Carter: In terms of control, the Dyvi could be simply deployed in these workflows. The control communication between control surface and the PM is via TCP/IP.
In terms of video being transported, this would be again brought back home via IP compressed or uncompressed and into the facility’s infrastructure. These sources would be received as any input and then is available to be routed anywhere within the Dyvi Production Core and switched, processed or used in an effect.
With more production equipment being deployed in IP and supporting a wider IT approach, there is a future for a COTS production switcher approach in any datacenter.
The Dyvi has the unique inter-PM link that allows all PMs to share sources and outputs. This inter-PM link allows the PM’s to be located in different physical locations. As an example, a Dyvi production core can consist of three PMs. Two PMs can be in the technical switching area and the third PM could be used for driving an in-vision screen in the studio set design.
The Broadcast Bridge: What new features will be demonstrated at the IBC Show in Amsterdam?
Carter: At this year’s IBC Show the Dyvi will be shown working on an IP network mixing HD into UHD, along with the ability to complete multi-layered composites that would take several M/Es to accomplish.
The Broadcast Bridge: Is this virtualized approach the future of production switchers? Why?
Carter: With more production equipment being deployed in IP and supporting a wider IT approach, there is a future for a COTS production switcher approach in any datacenter; while the production gallery stays consistent with the collaborative production team workflow being deployed.
The industry has witnessed central SDI routing being replaced by IP fabric with the associated benefits. The next natural step is embracing software approaches on networked COTS equipment and the models they potentially bring and access they support.
Combining the technology that the Dyvi is built on with the ease of creating great productions in a layering approach while deploying this distributed architecture with the ability to scale the production core gives more shared power at a lower cost.
Of course, we as design engineers also remain mindful of what’s important to the TDs and how they can do their job faster and more efficiently.
You might also like...
Brazil Adopts ATSC 3.0 For NextGen TV Physical Layer
The decision by Brazil’s SBTVD Forum to recommend ATSC 3.0 as the physical layer of its TV 3.0 standard after field testing is a particular blow to Japan’s ISDB-T, because that was the incumbent digital terrestrial platform in the country. C…
Designing IP Broadcast Systems: System Monitoring
Monitoring is at the core of any broadcast facility, but as IP continues to play a more important role, the need to progress beyond video and audio signal monitoring is becoming increasingly important.
Broadcasting Innovations At Paris 2024 Olympic Games
France Télévisions was the standout video service performer at the 2024 Paris Summer Olympics, with a collection of technical deployments that secured the EBU’s Excellence in Media Award for innovations enabled by application of cloud-based IP production.
HDR & WCG For Broadcast - Expanding Acquisition Capabilities With HDR & WCG
HDR & WCG do present new requirements for vision engineers, but the fundamental principles described here remain familiar and easily manageable.
What Does Hybrid Really Mean?
In this article we discuss the philosophy of hybrid systems, where assets, software and compute resource are located across on-prem, cloud and hybrid infrastructure.