Video: IPMX Makes Networks Easy

IPMX is bringing a standards-based, software deployable connectivity solution to ProAV. It stands itself in contrast to hardware-based IP technologies such as Zeevee and SDVoE which both aim to create de facto standards by building large alliances of companies based on their chips. IPMX’s aim is to open up the market to a free-to-access technology that can be implemented in hardware and software alike. In this way vendors have more freedom in implementation with the hope of wider interoperability, depending on IPMX adoption. These are amongst some of the business reasons behind IPMX which are covered in this talk Matrox’s David Chiappini.

In today’s video, Matrox’s Jean Lapierre looks at the technical side of IPMX to answer some of the questions from those who have been following its progress. AIMS, the Alliance for IP Media Solutions, are upfront about the fact that IPMX is a work in progress with important parts of the project dealing with carriage of HDCP and USB still being worked on. However, much has already been agreed so it makes sense to start thinking about how this would work in real life when deployed. For a primer on the technical details of IPMX, check out this video from Andreas Hildebrand.

 

 

Jean starts by outlining the aims of the talk; to answer questions such as whether IPMX requires a new network, expensive switches and PTP. IPMX, he continues, is a collection of standards and specifications which enable transport of HD, 4K or 8K video in either an uncompressed form or lightly compressed, visually lossless form with a latency of <1ms. Because you can choose to enable compression, IPMX is compatible with 1GB, CAT5e networks as well as multi-gigabit infrastructure. Moreover, there’s nothing to stop mixing compressed and uncompressed signals on the same network. In fact, the technology is apt for carrying many streams as all media (also known as ‘essences’ to include metadata) is sent separately which can lead to hundreds of separate streams on the network. The benefit of splitting everything up is that in the past if you wanted to read subtitles, you would have to decode a 3Gbps signal to access a data stream better measured in bytes per second. Receiving just the data you need allows servers or hardware chips to minimise costs.

Jean explains how multicast is used to deliver streams to multiple receivers and how receivers can subscribe to multiple streams. A lot of the time, video streams are used separately such as from a computer to a projector meaning exact timing isn’t needed. Even coming into a vision mixer/board doesn’t always need to be synchronised because for many situations, having a frame synchroniser on all inputs works well. There are, however, times when frame-accurate sync is important and for those times, PTP can be used. PTP stands for the Precise Time Protocol and if you’re unfamiliar, you can find out more here.

The upshot of using PTP with IPMX is that you can unlock perfect synchronisation for applications like video walls; any time you need to mix signals really. IMPX relaxes some of the rules of PTP that SMPTE’s ST 2059 employs to reduce the load on the grandmaster clocks. PTP is a very accurate timing mechanism but it’s fundamentally different from black and burst because it’s a two-way technology that relies on an ongoing dialogue between the devices and the clock. This is why Jean says that for anything more than a small network, you are likely to need a switch that is PTP aware and can answer the queries which would normally go to the single, central switch. In summary then, Jean explains that for many IPMX implementations you don’t need a new network, a PTP grandmaster or PTP aware switches. But for those wanting to mix signals with perfect sync or those who have a large network, new investment would reap benefits.

Watch now!
Speaker

Jean Lapierre Jean Lapierre
Senior Director, Advanced Technolgies
Matrox

Video: JT-NM ProAV Technology Roadmap for IPMX – Panel Discussion

Building on our coverage of IPMX to date, we see that this push to create a standard for IP for the ProAV market has been growing in momentum. With activity now in AIMS, AMWA, VSF and SMPTE, it’s important to bring together the thinking and have a central strategy which is why the JT-NM have released a roadmap. This has started by defining what is meant by ProAV: “The market for audiovisual (AV) communication equipment used in professional, industrial, commercial, and retail environments as a means to communicate with people.” As is noted by today’s panel, this is a wide definition and helps us understand why this is such a different proposition compared to the related ST 2110 and NMOS work for the broadcast market. There are lots of silos in the Pro AV space with many solutions being developed to cater to just one or two. This makes requirements capture difficult and has led to the fragmentation seen to date and partly why strong manufactures tend to be the ones pushing the market in a certain direction in contrast to the broadcast market where strong, early adopters set the direction for vendors.

 

 

The roadmap itself sets the aims of IPMX, of instance that is secure from the start, it will scale and integrate with 2110/AMWA broadcast installations and be able to be a software only solution. Phase 1 of the roadmap identifies existing standards and specifications which underpin the three IPMX tenents of security, Media, Control. Identified are NMOS IS-10 for access authentication and encryption, relevant 2110 standards, NMOS IS-04,-05, -07 and capabilities to use EDIDs. Phase 2 then adds HDCP support, support for ProAV audio formats, enhanced control such as for audio mapping (IS-08), legacy camera control via RS-232 etc. and then phase 3 will bring in media compression for WAN links, error correction techniques and closed captioning & subtitling. For control, it will add USB HID and a training and certification scheme will be launched.

The panel concludes discussing how IPMX is very much at home with live production which, of course, should help it dovetail well into the broadcast space. IPMX is seen by the panel to simplify the implementation of a 2110-like infrastructure which should allow easier and quicker installations than 2110 which are seen as larger, higher risk projects. IPMX could, it’s suggested, be used as an initial step into IP for broadcasters who seek to understand what they need to do organizationally and technically to adopt IP ahead of perhaps developing 2110 systems. But the technology is seen as going both ways allowing broadcasters to more readily adopt compressed workflows (whether JPEG XS or otherwise) and allow Pro AV players to bring uncompressed workflows more easily into their productions for those that would benefit.

Watch now!
Speakers

Karl Paulsen Karl Paulsen
CTO,
Diversified
Andrew Starks Andrew Starks
Director of Product Management,
Macnica America’s Inc.

David Chiappini David Chiappini
Chair, Pro AV Working Group, AIMS
Executive Vice President, Research & Development,
Matrox Graphics Inc.

Richard M. Friedel Richard Friedel
Executive Vice President, Technology & Broadcast Strategy,
21st Century Fox

Video: IPMX for Broadcast Installations?

IPMX, the new ProAV IP challenger spec, is taking shape promising to tame SMPTE’s ST 2110 standards, make PTP useable and extend AMWA into managing HDCP. Is this a tall order and can it actually deliver? Taking us through the ins and out is Jean Lapierre from Matrox.

With or without IPMX, ProAV is moving to IP whether with SDVoE, ZeeVee or something else. There are a number of competing technologies, but we hear from Jean that IPMX is the only software-defined one. This is important because if you don’t require a chip to be an IPMX product and participate in ProAV workflows, then anything can support IPMX such as PCs, Laptops and mobile phones.

 

 

IPMX based on RTP, ST 2110, ST 2059 PTP and AMWA specifications IS-04, IS-05, IS-08 (audio channel mapping), IS-11 for EDID handling as well as NMOS security and best practice guidance. This seems like a lot, but to cover media transfer, registration, control, security and interfacing with display screens, this is the range of tech needed.

Compared to SMPTE ST 2110, the PTP profile is easier to deploy and produces less traffic, explains Jean, and IPMX even works without PTP which support for asynchronous signals. Support of HDCO is included along with a lower-latency FEC mode for those that find 2022-7 too costly or impractical to deploy. Lastly, Jean points out that thanks to the in-built support for JPEG XS, IPMX can support UHD workflows within a 1GbE infrastructure.

Jean continues by discussing the compatibility between 2110 and IPMX. In principle IPMX and 2110 senders and receivers are interchangeable. Jean goes into more detail, but the example would be that IPMX is managing the HDCP encryption of the source using AMWA NMOS IS-11. IS-11 is, naturally available to be used with any other technology including ST 2110. If it’s adopted, then HDCP-protected material can flow between the two systems.

Watch now!
Speaker

Jean Lapierre Jean Lapierre
Senior Director, Advanced Technologies,
Matrox

Video: What is IPMX? – The IPMX Stack

“The AV over IP market has really matured [giving us] great quality, low latency and the kind of stability and features that customers are looking for,” says Andrew Starks from Macnica Technology. If that’s the case, why do we need another standard by the name of IPMX? Intended to open up the AV-over-IP market and provide customers with a better deal, Andrew takes us through the motivations of AIMS, AMWA, VSF, SMPTE and the other organisations involved.

IPMX is a set of open standards and specifications which seek to bring a technology platform to the Pro AV industry on which all vendors can interoperate and innovate. Built on SMPTE’s ST 2110 suite of standards and the accompanying NMOS APIs from AMWA, IPMX adds essential capabilities such as HDMI, HDCP and USB support to create a complete and reliable foundation for AV events and installations.

 

 

Whilst there are a number of successful AV initiatives such as SDVoE, these are typically alliances built around a single-vendor hardware solution which is available to vendors in the alliance. This provides interoperability within that ecosystem but, explains Andrew, it prevents wider interoperability between vendors of different alliances. It also makes it hard to any vendor to innovate in the core feature set since that’s delivered from the single source relegating innovation to ‘plumbing’. For the vendors, at best, this means they have to contend with multiple, incompatible product lines and complicated support. Overall this results in a bad end user experience as they operate multiple islands which can have conflicting network requirements, i.e. 10GbE vs 1GbE.

IPMX can be implemented in software as well as hardware using compressed or uncompressed video with a focus on fully featured discovery as this has been identified as being as important as the ability to carry video. Timing has been made flexible such that it can operate with or without PTP which is one of a number of ways that it’s anticipated IPMX will be able to merge in with ST 2110 infrastructures.

Andrew finishes off his talk with a look at the tech stack of IPMX with layer 2 options from 1 to 100GbE connections supported on which RTP and PTP run. SMPTE’s ST 2110 standards feature heavily alongside a new standard for HDCP in 2110, a VSF spec for FEC and new specifications from AMWA for asynchronous control traffic like EDID, Serial, CEC, USB etc. Finally, there are the main APIs such as IS-04, -05 etc. as well as the application layer which uses OAuth2 for authenticating and has an RDS server for discovery. Lastly, there is a look at the JT-NM roadmap to see how the IPMX work will continue to advance throughout this year.

Watch now!
Speakers

Andrew Starks Andrew Starks
Director of Product Management,
Macnica America’s Inc.