{"id":15222,"date":"2026-07-24T09:24:55","date_gmt":"2026-07-24T01:24:55","guid":{"rendered":"https:\/\/telycam.com\/?p=15222"},"modified":"2026-07-24T09:27:52","modified_gmt":"2026-07-24T01:27:52","slug":"what-is-ndi-high-bandwidth","status":"publish","type":"post","link":"https:\/\/telycam.com\/what-is-ndi-high-bandwidth.html","title":{"rendered":"What Is NDI High Bandwidth?"},"content":{"rendered":"\t\t
\n\t\t\t\t
\n \t\t\t

What Is NDI High Bandwidth?<\/h1>\n\t\t\t\t\t

The Standard for Broadcast-Quality IP Video<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

NDI made it possible to move video through standard IP networks instead of traditional SDI cables. But as production demands increased \u2014 higher resolutions, lower latency, and broadcast-quality workflows \u2014 another question emerged: how do you send the highest-quality video over IP without compromise?<\/strong>

That's where NDI High Bandwidth<\/strong> comes in.

Designed for professional live production, NDI High Bandwidth prioritizes image quality and real-time performance, delivering visually lossless video with extremely low latency over standard network infrastructure. In this guide, we'll explain what NDI High Bandwidth is, how it differs from other NDI formats, and where it fits in modern broadcast and AV workflows.<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

What Is NDI? A Quick Foundation<\/h2>\n\t\t\t\t\t

NDI \u2014 short for Network Device Interface \u2014 is a software protocol developed by NewTek that enables video, audio, and metadata to be transmitted and received over a standard local area network.

Before NDI, getting video between devices meant dedicated point-to-point connections: one SDI cable from camera to switcher, another from switcher to monitor, another from replay to graphics. Every signal required its own physical path. NDI replaces that model with a shared network approach \u2014 devices discover each other automatically, and any source on the network can be received by any compatible device on the same network, without additional cabling or routing hardware.

NDI is not a single standard but a family of versions, each making a different trade-off between image quality, latency, and bandwidth consumption. Understanding which version is which matters in practice, because the differences between them are significant enough to affect production decisions.

The three versions in current use are:<\/p>\n\t\t

\n
\n \t\t\t
<\/path><\/svg><\/div>\t\t\t\t\t

FULL SPEED<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

NDI High Bandwidth<\/h3>\n\t\t\t\t\t

The highest-quality version of the NDI protocol. Using intra-frame SpeedHQ compression, it delivers visually lossless video with sub-frame latency \u2014 typically less than 16ms at 1080p60. Bandwidth requirements scale with resolution: 100\u2013150 Mbps for 1080p60, rising to 250 Mbps or more for 4K60.<\/p>\n\t\t

\n \t\t\t
<\/path><\/svg><\/div>\t\t\t\t\t

Up to 150 Mbps<\/strong><\/p>\n\t\t\t\t\t

<\/path><\/svg><\/div>\t\t\t\t\t

Sub-frame Latency<\/strong><\/p>\n\t\t\n<\/div>\n\n<\/div>\n\n<\/div>\n

\n
\n \t\t\t
<\/path><\/svg><\/div>\t\t\t\t\t

HX3 HYBRID<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

NDI|HX3<\/h3>\n\t\t\t\t\t

The practical sweet spot for most modern production environments. It uses short-GOP H.264 or H.265 compression to reduce bandwidth to around 8\u201340 Mbps at 1080p60 \u2014 a fraction of NDI High Bandwidth \u2014 while maintaining latency low enough for real-time live switching. For multi-camera deployments on standard gigabit networks, HX3 is typically the most efficient choice.<\/p>\n\t\t

\n \t\t\t
<\/path><\/svg><\/div>\t\t\t\t\t

8\u201340 Mbps<\/strong><\/p>\n\t\t\t\t\t

<\/path><\/svg><\/div>\t\t\t\t\t

Live Switching Ready<\/strong><\/p>\n\t\t\n<\/div>\n\n<\/div>\n\n<\/div>\n

\n
\n \t\t\t
<\/path><\/svg><\/div>\t\t\t\t\t

HX2 REMOTE<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

NDI|HX2<\/h3>\n\t\t\t\t\t

Optimized for bandwidth-constrained environments \u2014 remote production, campus distribution, and networks where video shares infrastructure with other traffic. Using long-GOP H.265, it achieves bitrates as low as 2\u201316 Mbps, low enough to run reliably over standard Wi-Fi. The tradeoff is latency: HX2 typically introduces 100\u2013300ms of delay, which makes it unsuitable for live switching but practical for monitoring feeds, secondary distribution, and recording.<\/p>\n\t\t

\n \t\t\t
<\/path><\/svg><\/div>\t\t\t\t\t

2\u201316 Mbps<\/strong><\/p>\n\t\t\t\t\t

<\/path><\/svg><\/div>\t\t\t\t\t

Remote Distribution<\/strong><\/p>\n\t\t\n<\/div>\n\n<\/div>\n\n<\/div>\n\t\t\t

This guide focuses on NDI High Bandwidth \u2014 the version designed for professional live production where image quality and switching performance are the priority.<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

What Is NDI High Bandwidth?<\/h2>\n\t\t\t\t\t

NDI High Bandwidth \u2014 also referred to as Full NDI \u2014 is the version of the NDI protocol that prioritizes image quality and low latency above all else, transmitting broadcast-grade video over a standard ethernet network without the quality compromises that higher compression introduces.<\/strong>

Think of it as the FLAC of video transmission. Just as FLAC preserves the full audio signal with minimal compression while MP3 discards detail to reduce file size, NDI High Bandwidth preserves virtually all of the original image data while using only the compression necessary to make network transmission practical. The result is video that is functionally indistinguishable from the uncompressed SDI <\/u><\/strong><\/a>signal coming out of the same camera.

The codec that makes this possible is SpeedHQ<\/strong> \u2014 an intra-frame compression format where every frame is compressed independently, with no reference to surrounding frames. Because each frame contains its complete image data, a receiver can decode and display it immediately without waiting for additional reference information. This is what enables NDI High Bandwidth's defining performance characteristic: latency of less than one frame \u2014 typically under 16ms at 1080p60 \u2014 fast enough to be effectively invisible in a live production context.<\/p>\n\t\t

\n
\n \t\t\t

LATENCY<\/p>\n\t\t

\n \t\t\t

<16ms<\/strong><\/p>\n\t\t\n<\/div>\n\t\t\t

Sub-frame end-to-end<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

TYPICAL BITRATE<\/p>\n\t\t

\n \t\t\t

150<\/strong><\/p>\n\t\t\t\t\t

Mbps<\/strong><\/p>\n\t\t\n<\/div>\n\t\t\t

1080p60 Stream<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

COLOR DEPTH<\/p>\n\t\t

\n \t\t\t

10<\/strong><\/p>\n\t\t\t\t\t

bit<\/strong>
<\/p>\n\t\t\n<\/div>\n\t\t\t

4:2:2 Color Support<\/p>\n\t\t\n<\/div>\n\n<\/div>\n\t\t\t

Visually Lossless Image Quality<\/h3>\n\t\t\t\t\t

NDI High Bandwidth operates at 100\u2013150 Mbps for 1080p60, scaling to 250 Mbps or more for 4K60. To put that in context, a standard streaming video might use 5\u201315 Mbps; NDI High Bandwidth carries ten to fifty times that data rate, preserving color accuracy, edge sharpness, and fine detail at a level that is indistinguishable from uncompressed SDI to the human eye. For productions where image quality is non-negotiable \u2014 broadcast output, large-venue IMAG screens, color-critical monitoring \u2014 this is the relevant standard.<\/p>\n\t\t\t\t\t

Sub-Frame Latency<\/h3>\n\t\t\t\t\t

The latency of an NDI High Bandwidth stream at 1080p60 is typically under 16ms \u2014 less than a single frame at that frame rate. In practice, this means that what the camera sees and what appears on the production monitor or switcher output are effectively simultaneous. For live applications where timing precision matters \u2014 multi-camera switching, large-screen audience display synchronized with live audio, or fast-motion sports replay \u2014 this level of latency is what separates a system that feels immediate from one that feels slightly behind.<\/p>\n\t\t\n<\/div>\n

\n \t\t\t

Why \"Intra-Frame\" Matters:
The Secret Behind Low Latency<\/h2>\n\t\t\t\t\t

NDI High Bandwidth's low latency isn't a marketing claim \u2014 it's a direct consequence of how the video is compressed. Understanding the difference between intra-frame and inter-frame compression makes it clear why some video formats can be switched live without hesitation, and others can't.<\/p>\n\t\t\t\t\t

Two Ways to Compress Video<\/h3>\n\t\t\t\t\t

Every video compression system faces the same fundamental challenge: raw video contains an enormous amount of data, and transmitting or storing all of it is expensive. The solution is to find and remove redundancy \u2014 information that doesn't need to be stored explicitly because it can be inferred from something else.

There are two fundamentally different approaches to doing this:

Inter-frame compression<\/strong> looks for redundancy across multiple frames. If a background stays the same from one frame to the next, there's no need to store it again \u2014 only the parts that changed need to be recorded. This is how most consumer video formats work: H.264 and H.265, the codecs used in streaming services, YouTube, and most cameras, all use inter-frame compression. The efficiency gains are significant \u2014 bitrates drop dramatically \u2014 but there's a structural cost. To decode any given frame, the decoder may need information from frames that came before it, and sometimes frames that come after it. It has to wait until it has enough reference data before it can produce a complete image.<\/p>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t
\r\n \r\n \r\n
\r\n
\r\n \r\n <\/path>\r\n <\/path>\r\n <\/path>\r\n <\/svg>\r\n

Inter-frame<\/h2>\r\n <\/div>\r\n

\r\n Commonly used in H.264\/H.265 (HEVC), this method looks for changes between multiple frames.\r\n <\/p>\r\n \r\n \r\n

\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
\r\n I<\/span>\r\n <\/div>\r\n
<\/div>\r\n
<\/div>\r\n <\/div>\r\n
Requires Frame Buffer<\/div>\r\n <\/div>\r\n \r\n \r\n
    \r\n
  • 01.<\/span> Smaller file sizes for streaming.<\/li>\r\n
  • 02.<\/span> High CPU load for encoding.<\/span><\/li>\r\n
  • 03.<\/span> Significant Latency (100ms+).<\/span><\/li>\r\n <\/ul>\r\n <\/div>\r\n\r\n \r\n
    \r\n
    \r\n \r\n <\/path>\r\n <\/polyline>\r\n <\/svg>\r\n

    Intra-frame<\/h2>\r\n <\/div>\r\n

    \r\n The NDI High Bandwidth \/ SpeedHQ standard. Every single frame is its own complete picture.\r\n <\/p>\r\n \r\n \r\n

    \r\n
    \r\n
    \r\n <\/polyline><\/svg>\r\n <\/div>\r\n
    \r\n <\/polyline><\/svg>\r\n <\/div>\r\n
    \r\n <\/polyline><\/svg>\r\n <\/div>\r\n <\/div>\r\n
    Independent Frames<\/div>\r\n <\/div>\r\n \r\n \r\n
      \r\n
    • 01.<\/span> Ultra-low CPU overhead.<\/li>\r\n
    • 02.<\/span> Instant frame access.<\/li>\r\n
    • 03.<\/span> Near-Zero Latency (<16ms).<\/strong><\/li>\r\n <\/ul>\r\n <\/div>\r\n\r\n<\/div>\r\n\r\n