This year, as I wandered about the 2024 NAB Show, sending video (VoIP) or audio (AoIP) over the Internet for remote production or editing seemed on display everywhere.
However, I was buried in acronyms: NDI 6, SMPTE 2110, RIST, Dante… each more confusing than the last. Time for me to do some homework to figure this out. Here’s what I learned.
Traditionally, when editing media – especially video – we edited on a personal computer, using storage that was directly attached to it. Then, as editing expanded into workgroups, we connected our computers to fast, network-attached storage using a local-area network.
Today, as more productions are edited by editors scattered across the globe, we need to find a way to have editors access media stored remotely, but monitored and edited locally. These connections need to be fast, reliable, high-quality and easy for non-IT people to create.
This need gave rise to a variety of internet transport protocols to move media from a remote server to a local editor with a minimum of delay (latency) and maximum of quality. Currently, the most popular transport formats are:
NOTE: There are others: SMPTE 2022, ASPEN, and NMI from Sony.
All these protocols allow an editor to connect their computer to the Internet to find, view, access and edit media using an Ethernet connection. For example, one or more cameras can be attached to the web, then viewed from multiple locations without requiring the traditional infrastructure of a control room or video village.
As you might suspect, these protocols are not compatible with each other and none of them dominate the market. Which means that you could run into any of these as you move between projects.
NOTE: These protocols are separate from the Camera-to-Cloud workflows from Frame.io or Blackmagic Design.
Here’s what these key standards are and how they are used.
Image courtesy of Pixabay, via Pexels.com
NDI
NDI (Network Device Interface) was developed by NewTek (now a part of VizRT) and first released in 2015. It was designed to run over gigabit Ethernet, using the NDI codec. This software specification enabled high-definition video to be delivered, received, and communicated in a low-latency, high-quality manner.
Devices connected to the web via NDI can be automatically discovered by other devices using Bonjour, NDI Access (a manual process) or Discovery Server. The NDI specification is royalty-free, works bi-directionally, supports up to 4K with unlimited audio channels, and allows for frame-accurate switching, making it useful for both post-production and live productions. It communicates using TCP, Multi-TCP and UDP.
This spec supports video, audio, metadata, and tally lights. Devices engineered to the NDI spec are plug-and-play interoperable and require minimal network knowledge.
The latest version is NDI 6. This adds support for HDR encoding and playback, native 10-bit color depth support, and more robust support for SDI.
Companies providing NDI support include: Microsoft Skype, EVS, Avid, VizRT, NewTek, XSplit, Chyron and more.
SMPTE 2110
SMPTE 2110 is a suite of standards from the Society of Motion Picture and Television Engineers (SMPTE) that describes how to send digital media over an IP network. It is intended to be used within broadcast production and distribution facilities where quality and flexibility are more important than bandwidth efficiency.
NOTE: NDI runs on a gigabit (1G) network. SMPTE-2110 requires 10-gigabit (10G) speeds.
This spec also supports video, audio, metadata, and tally lights.
This standard was first released in November, 2017. While ST-2110-20 supports uncompressed video transport, the majority of applications use ST-2110-22, which transports using compressed video. The recently announced cloud features for Blackmagic Design support SMPTE-2110.
The majority of SMPTE 2110-22 implementations use the JPEG XS lightweight low latency compression standard created by the Joint Photographic Experts Group to answer the requirements of the SMPTE 2110 standard. The JPEG XS codec has only a few video lines of latency in software and hardware implementations, which is less than 1 millisecond.
RIST
RIST (Reliable Internet Stream Transport) is an open-source, open specification transport protocol designed for reliable transmission of video and audio over lossy networks (including the internet) with low latency and high quality.
Released in 2018, it was developed and promoted by the RIST Forum, a not-for-profit association of leading vendors and global broadcast operators. By adhering to established protocols like RTP and SMPTE-2022, RIST provides a standardized framework for efficient RIST networking management.
RIST provides secure connections, network resilience in case a portion of a connection breaks, multicast which supports sending the same signal to multiple locations at once, and the ability to throttle bandwidth if a connection deteriorates with minimal quality loss.
While RIST can be used in multiple applications, it is principally used in a broadcast or streaming environment for distribution.
DANTE
Dante (Digital Audio Network Through Ethernet) was developed by Audinate in 2006. It is the product name for a combination of software, hardware, and network protocols that delivers uncompressed, multi-channel, low-latency digital audio over a standard Ethernet network at speeds as slow as 100 mbps.
While it is ideally suited to recording studios and post-production, it is also used in live sound, houses of worship, education, commercial sound installations like stadiums or airports), and for broadcast. Most often, it is used in applications where a large number of audio channels must be transmitted over relatively long distances or to multiple locations.
Signals are carried as uncompressed audio at 44.1kHz/48kHz/96kHz, with 16-, 24-, or 32-bit resolution. They have a dynamic range that exceeds 100dB and total harmonic distortion under 0.01% at +4dBu.
Dante requires Dante Controller, a cross-platform, free utility from Audinate. This provides device discovery, signal routing, network status and bandwidth. While all connected devices must use Dante, Audinate provides adapters which convert analog audio into Dante-compatible digital audio. This allows connecting mixers, keyboards, mic preamps, direct boxes, analog signal processors, and more to a Dante network. Currently, Audinate has over 350 licensees who have produced over 3,000 Dante-compatible products.
Any current or legacy audio device can be connected to a Dante network.
SUMMARY
There is no doubt the need to quickly and accurately move media around the web while maintaining the highest quality will continue to grow. The challenge now is to make sure that your gear will support the formats you need. Not all these protocols are compatible and not all are designed for post-production.
ADDITIONAL RESOURCES
The following websites were used in writing this article:
NDI
SMPTE-2110
Dante
RIST
4 Responses to Why Should We Care about NDI, SMPTE 2110 or RIST? They Are Ways to Move Media Remotely.
ST2110 will shake up remote production. Take F1 for example. The circuits know where broadcasters put cameras so they’ll put a bunch of them on posts around the track – and possibly in Blimps or Drones.
Instead of sending OB trucks to the race, broadcasters will hire the number of cameras they need for good coverage. The ST2110 protocol will allow complete lineup and PZT control of the cameras from anywhere on the Internet.
AIs can be trained to know what a TV programme about F1 races should look like, will vision mix the show, and add archive data as needed. Vision analysis and AI VO will soon be good enough to provide a running commentary. AI will also be able to produce real-time and post-event highlights for TV and Social platforms
This is coming. It will be much cheaper. Clearly, this way of working will appeal to broadcasters for covering a variety of sports and – even coronations. Watch out for the crowning of King William V!
Simon:
Yup. I agree and this vision is both exciting (for the broadcaster) and terrifying (for the crew that used to be paid to produce these events).
ST-2110 can cary video, audio, intercom, tally lights, and metadata. Totally redefines remote production.
Larry
To me, the analogy is like :
NDI –> MP3 – lower barrier of entry to implement, cheaper than ST2110, suitable for general public / smaller medium needs, quality : very good.
ST2110 –> WAV/AAC – higher barrier of entry, more costly, good for broadcaster/enterprise, quality : exceptional.
is this true ?
Irsan:
This is an interesting analogy, but I’m not sure it’s true.
I’ll try to reach the folks at VizRT, who own NDI, to get their reaction.
Larry