Keycode Media

Video: A Cloudy Future For Post Production

Posted on by Russell Trafford-Jones

Even before the pandemic, post-production was moving into the cloud. With the mantra of bringing your apps to the media, remote working was coming to offices and now it’s also coming to homes. As with any major shift in an industry, it will suit some people earlier than others so while we’re in this transition, it’s work taking some time to ask why people are doing this, why some people are not and what problems are still left unsolved. For a wider context on the move to remote editing, watch this video from IET Media.

This video sees Key Code Media CTO,Jeff Sengpiehl talking to Bebop Technology’s Michael Kammes, Ian McPherson from AWS and Ian Main from Teradici. After laying the context for the discussion, he asks the panel why consumer cloud solutions aren’t suitable for professionals. Michael picks this up first by picking on consumer screen sharing solutions which are optimised for getting a task done and they don’t offer the fidelity and consistency you need for media workflows. When it comes to storage at the consumer level, the cost usually prevents investment in the hardware which would give the low-latency, high-capacity storage which is needed for many professional video formats. Ian then adds that security plays a much bigger role in professional workflows and the moment you bring down some assets to your PC, you’re extending the security boundary into your consumer software and to your house.

 

 

The security topic features heavily in this conversation and Michael talks about the Trusted Partner Network who are working on a security specification which, it is hoped, will be a ‘standard’ everyone can work to in order to show a product or solution is secure. The aim is to stop every company having their own thick document detailing their security needs because that means each vendor has to certify themselves time and time again against similar demands but which are all articulated differently and therefore defended differently. Ian explains that cloud providers like AWS provide better physical security than most companies could manage and offer security tools for customers to secure their solution. Many are hoping to form their workflows around the Movielabs 2030 vision which recommends ways to move content through the supply chain with security and auditing in mind.

“What’s stopping people from adopting the cloud for post-production?”, poses Jeff. Cost is one reason people are adopting the cloud and one reason others aren’t. Not dissimilar to the ‘IP’ question in other parts of the supply chain, at this point in the technology’s maturity, the cost savings are most tangible to bigger companies or those with particularly high demands for flexibility and scalability. For a smaller operation, there may simply not be enough benefit to justify the move. Particularly as it would mean adopting tools that take time to learn so, even if temporary, slow down an editor’s ability to deliver a project in the time they’re used to. But on top of that, there’s the issue of cost uncertainty. It’s easy to say how much storage will cost in the cloud, but when you’re using dynamic amounts of computation and moving data in and out of the cloud, estimating your costs becomes difficult and in a conservative industry this uncertainty can form part of a blocker to adoption.

Starting to take questions from the audience, Ian outlines some of the ways to get terabytes of media quickly into the cloud whilst Michael explains his approach to editing with proxies to at least get you started or even for the whole process. Conforming to local, hi-res media may still make sense out of the cloud or you have time to upload the files whilst the project is underway. There’s a brief discussion on the rise of availability of Macs for cloud workflows and a discussion about the difficulty, but possibility, of still having a high-quality monitoring feed on a good monitor even if your workstation is totally remote.

Watch now!
Speakers

Ian Main Ian Main
Technical Marketing Principal,
Teradici Corporation
Ian McPherson Ian McPherson
Head of Global Business Development,
Media Supply Chain
Michael Kammes Michael Kammes
VP Marketing & Business Development,
BeBop Technology
Jeff Sengpiehl Jeff Sengpiehl
CTO
Key Code Media

Video: The State of NDI in 2021 & NDI Version 5

Posted on by Russell Trafford-Jones

Released in 2015, NDI is a royalty-free technology that implements low-latency video, audio, tally and control over gigabit networking. Since 2015, NDI has grown in popularity immensely and is now very widely used within AV, live streaming production and in cloud workflows. Developed by Newtek, now part of VizRT, it allows computers to easily push video from programs onto the local network whether from Teams or Skype, a video Editor, OBS or anything else. Many vendors have taken the NDI binaries and integrated them into their products.

On The Broadcast Knowledge we’ve explored how NDI compliments SMPTE’s ST 2110 suite of standards which primarly help move uncompressed payloads around a broadcast suite. In this panel put on by the IET at IBC we explored the benefits of 2110 and NDI to understand where the overlap lies. And for a deeper dive into NDI including some demos, we also featured this talk by SMPTE and VizRT.

 

 

In today’s video from Key Code Media, we hear from Newtek’s Dr. Andrew Cross, creator of NDI on what’s new in NDI’s latest release, 5.0. Jeff Sengpiehl from Key Code Media explains that NDI 5.0 brings with it improvements in multi-platform support including native support for Apple Silicon. 5.0 also includes plugins for any program to share its audio over NDI as well as a way of sharing a link to get someone to share their video and audio by NDI. Part of the big changes, though in this latest version is the addition of ‘reliable UDP’ and ‘NDI Bridge’. Based on Google’s QUIC this provides a way of sending NDI over lossy networks allowing for it to recover lost data and deal with network congestion. This ties in nicely with ‘NDI Bridge’ which allows two or three separate locations to share their NDI sources and destinations seamlessly. These additional features bring NDI outside of the LAN. Being a layer 2 protocol, it’s always been seen as a local protocol even when deployed in the cloud.

The majority of the video features Dr. Cross answering questions from Jeff and viewers. These questions include:
What are the pain points with audio? Is NDI Audio Direct a replacement for Dante? Maintaining synchronisation in multi-location systems. The significance of support for ARM chips, bidirectional use of NDI, NDI Bridge security, 10-bit colour support, NDI’s place in the ProAV market and the possibility of NDI becoming open source or a standard.

Watch now!
Speakers

Andrew Cross Andrew Cross
Creator of NDI & President of Global Research and Development
VizRT (NewTek/NDI)
Jeff Sengpiehl Moderator: Jeff Sengpiehl
Chief Technologist
Key Code Media

Video: How to Deploy an IP-Based Infrastructure

Posted on by Russell Trafford-Jones

An industry-wide move to any new technology takes time and there is a steady flow of people new to the technology. This video is a launchpad for anyone just coming into IP infrastructures whether because their company is starting or completing an IP project or because people are starting to ask the question “Should we go IP too?”.

Keycode Media’s Steve Dupaix starts with an overview of how SMPTE’s suite of standards called ST 2110 differs from other IP-based video and audio technologies such as NDI, SRT, RIST and Dante. The key takeaways are that NDI provides compressed video with a low delay of around 100ms with a suite of free tools to help you get started. SRT and RIST are similar technologies that are usually used to get AVC or HEVC video from A to B getting around packet loss, something that NDI and ST 2110 don’t protect for without FEC. This is because SRT and RIST are aimed at moving data over lossy networks like the internet. Find out more about SRT in this SMPTE video. For more on NDI, this video from SMPTE and VizRT gives the detail.

 

 

ST 2110’s purpose is to get high quality, usually lossless, video and audio around a local area network originally being envisaged as a way of displacing baseband SDI and was specced to work flawlessly in live production such as a studio. It brings with it some advantages such as separating the essences i.e. video, audio, timing and ancillary data are separate streams. It also brings the promise of higher density for routing operations, lower-cost infrastructure since the routers and switches are standard IT products and increased flexibility due to the much-reduced need to move/add cables.

Robert Erickson from Grass Valley explains that they have worked hard to move all of their product lines to ‘native IP’ as they believe all workflows will move IP whether on-premise or in the cloud. The next step, he sees is enabling more workflows that move video in and out of the cloud and for that, they need to move to JPEG XS which can be carried in ST 2110-20. Thomas Edwards from AWS adds their perspective agreeing that customers are increasingly using JPEG XS for this purpose but within the cloud, they expect the new CDI which is a specification for moving high-bandwidth traffic like 2110-20 streams of uncompressed video from point to point within the cloud.

John Mailhot from Imagine Communications is also the chair of the VSF activity group for ground-cloud-cloud-ground. This aims to harmonise the ways in which vendors provide movement of media, whatever bandwidth, into and out of the cloud as well as from point to point within. From the Imagine side, he says that ST 2110 is now embedded in all products but the key is to choose the most appropriate transport. In the cloud, CDI is often the most appropriate transport within AWS and he agrees that JPEG XS is the most appropriate for cloud<->ground operations.

The panel takes a moment to look at the way that the pandemic has impacted the use of video over IP. As we heard earlier this year, the New York Times had been waiting before their move to IP and the pandemic forced them to look at the market earlier than planned. When they looked, they found the products which they needed and moved to a full IP workflow. So this has been the theme and if anything has driven, and will continue to drive, innovation. The immediate need provided the motivation to consider new workflows and now that the workflow is IP, it’s quicker, cheaper and easier to test new variation. Thomas Edwards points out that many of the current workflows are heavily reliant on AVC or HEVC despite the desire to use JPEG XS for the broadcast content. For people at home, JPEG XS bandwidths aren’t practical but RIST with AVC works fine for most applications.

Interoperability between vendors has long been the focus of the industry for ST 2110 and, in John’s option, is now pretty reliable for inter-vendor essence exchanges. Recently the focus has been on doing the same with NMOS which both he and Robert report is working well from recent, multi-vendor projects they have been involved in. John’s interest is working out ways that the cloud and ground can find out about each other which isn’t a use case yet covered in AMWA’s NMOS IS-04.

The video ends with a Q&A covering the following:

  • Where to start in your transition to IP
  • What to look for in an ST 2110-capable switch
  • Multi-Level routing support
  • Using multicast in AWS
  • Whether IT equipment lifecycles conflict with Broadcast refresh cycles

    • Watch now!
    Speakers

    John Mailhot John Mailhot
    CTO & Director of Product Management, Infrastructure & Networking,
    Imagine Communications
    Ciro Noronha Ciro Noronha
    Executive Vice-President of Engineering,
    Cobalt Digital
    Thomas Edwards Thomas Edwards
    Principal Solutions Architect & Evangelist,
    Amazon Web Services
    Robert Erickson Robert Erickson
    Strategic Account Manager Sports and Venues,
    Grass Valley
    Steve Dupaix Steve Dupaix
    Senior Account Executive,
    Key Code Media