16-Channel I/O Board for handling Audio Data Embedded in a Digital Serial Video Stream corresponding to SMPTE 259M (SD), SMPTE 292M (HD) and SMPTE 424M/425M (3G), and for embedding and de-embedding SMPTE 2020-Compliant Metadata

  • I/O board for 16 audio channels
  • De-embeds, embeds, replaces, and/or deletes audio compliant with SMPTE 272M-AC (SD), SMPTE 299M (HD), and SMPTE 424M/425M (3G)
  • Provides de-embedding and embedding of 2 SMPTE 2020-1/-2-compliant metadata streams
  • Direct link to XDED Dolby decoder and XDEE Dolby encoder for de-embedding and embedding asynchronous Dolby E audio signals
  • Free channel assignment and routing on the NEXUS
  • Internal DSP for audio processing
  • Adjustable video delay: 0-15 frames (SD), 0-8 frames (HD/3G)
  • Adjustable delay (0-170 ms @ 48 kHz) for audio and metadata send channels
  • Down quantization of 20/24 bit source formats to 16 bit available, with dithering and noise shaping
  • Built-in test-pattern video generator
  • SMPTE RP165 (SD) compliant EDH check and insertion, CRC check and correction for HD/3G formats
  • Bypass mode; can also be used for breakdown protectionung
  • Configured with BNC or optical ports (LC, single mode)
  • Optional XSRCA 02 SRC module for all audio channels
  • Officially certified by Dolby® for Dolby E signal transmission

Dolby and the double-D symbol are registered trademarks of Dolby Laboratories.

 data sheet (275 KB)

In addition to SD and HD video, XHDI also supports HD-3G formats. The XHDI enables the NEXUS system to receive, add, and process audio that is integrated in a digital serial video stream complying with either of the SMPTE 259M (SD), SMPTE 292M (HD), and SMPTE 424M/425M (3G) standards. It allows for embedding and de-embedding metadata streams which are used particularly in conjunction with the transport and processing of multi-channel audio. The board provides an adjustable video delay for compensating latencies between video and audio. Moreover, it features a video input and a video output. This allows any video to be forwarded simply and without modification, with no need for additional hardware components. If only audio data and or metadata are to be extracted from a video stream, the XHDI can be used as an input board.
The XHDI is designed as a universal SD and HD/3G interface. It offers a multiplicity of possibilities for reading audio from a video data stream, or adding it to a video data stream:

De-Embedding  Any audio can be extracted from the video stream and be used and forwarded on the NEXUS audio network. Since the video stream remains unaffected, the output video signal’s audio data is retained.

Embedding  Any audio signals present on the NEXUS system can be embedded into the video data stream. The data blocks and channels to be written in the SDI signal are selectable freely. Other blocks audio data, which were contained in the video stream previously, are retained. Channels from the same block can be carried over as they were.

› Replacing  The replace mode allows individual audio streams to be replaced in non-standard data streams. In this mode, only the audio data is written to the existing data structure. This structure can be arranged in different ways.

Deletion  Audio blocks already contained in the video signal or marked for deletion can be deleted or overwritten entirely.

Metadata can be embedded and de-embedded. The board supports two metadata streams in each direction. Allocation and position is user definable. There is no replace mode for metadata – deletion of all metadata from a video stream is possible. Existing metadata can be overwritten.
A comprehensive software package enables all functions of the XHDI to be controlled from any PC.

High Flexible

The XHDI board handles signals compliant with the SMPTE 259M (4 : 2 : 2, Component, 270 Mbps), SMPTE 292M (4 : 2 : 2, Component, 1485/1483.5 Mbps), and SMPTE 425M (4 : 4 : 4 or 4 : 2 : 2 2970/2967 Mbps) standards. The standard format and any sub-formats under it are recognized automatically.

Supported video formats include:

Video Standard




SMPTE 425M 4:2:2 1920x1080 / 60p 3G yes
  1920x1080 / 50p   yes
SMPTE 425M 4:4:4 1920x1080 / 60i   -
  1920x1080 / 30p (PsF)   -
  1920x1080 / 30p   -
  1920x1080 / 50i   -
  1920x1080 / 25p (PsF)   -
  1920x1080 / 25p   -
  1920x1080 / 24p   -
  1280x720 / 60p   -
  1280x720 / 50p   -
  1280x720 / 25p   -
  1280x720 / 24p   -
SMPTE 296M 1280x720 / 60p HD yes
  1280x720 / 50p   yes
  1280x720 / 30p   -
  1280x720 / 25p   -
  1280x720 / 24p   -
SMPTE 274M 1920x1080 / 60i HD yes
  1920x1080 / 30p   yes
  1920x1080 / 50i   yes
  1920x1080 / 25p   yes
  1920x1080 / 25p   yes
  1920x1080 / 24p   yes
  1920x1080 / 30p (PsF)   -
  1920x1080 / 25p (PsF)   -
  1920x1080 / 24p (PsF)   -
SMPTE 260M 1920x1035 / 60i HD -
SMPTE 259M 1920x1080 / 50i HD -
SMPTE 260M 1440x487 / 60i
(or x507)
SD yes
  1440x576 / 50i   yes

PCM Audio Data

The SMPTE 272M-AC (SD) and SMPTE 299M (HD/3G) standards specify the transport of audio embedded in video streams. They define a maximum of four data blocks, each containing four 20/24-bit audio channels. Both standards are supported fully by the XHDI . It de-embeds and/or embeds single or multiple channels into the data stream at the same time.
The SMPTE-272M-AC (SD) standard supports various types of embedding audio data into the video stream. This can lead to problems in long transmission lines and may make further incremental embedding of audio data impossible. Therefore, the XHDI board allows for existing data structures to be rearranged completely and, if necessary, to be rewritten, including the newly added audio data. According to SMPTE 299M (HD/3G), there are no restrictions regarding embedding. Therefore, no compatibility issues are to be expected as a consequence of different embedding algorithms.
A FIFO buffer is used to reconstruct the necessarily isochronous audio stream. The buffer size can be changed to suit the received signal in case of problems.
In HD/3G mode, the XHDI board can offer a special function: It tries to achieve an identical latency for all blocks in the Embedding and De-Embedding modes (but not in the Replace mode). This functionality is useful in multi-channel applications where the maximum number of 4 channels per audio block is exceeded. This functionality requires that the structure of the existing embedded audio be identical to the new audio. If the function is used only for embedding (without de-embedding existing audio), the XHDI can guarantee an identical latency for all audio blocks.
If the audio data to be processed is asynchronous to the NEXUS system, an SRC (sample-rate converter) module can be installed. The converter synchronizes the 16 send channels as well as the 16 receive channels to the NEXUS system clock.
All send channels feature an adjustable audio delay for compensating latencies between video and audio. In addition, the board enables the video to be delayed by up to 15 frames (or up to 8 frames with HD/3G signals).
The NEXUS system can be synchronized to the video signal and, with asynchronous audio, to the audio word clock encoded in the data stream. (only HD/3G)
The (AES-3 compliant) ancillary data contained in the audio stream is made available for software evaluation. On the sending side, the ancillary data can either be transparently adopted from AES-3 compliant signal sources or be regenerated by the XCPU board.

Compressed Audio, SDI Link

SDI signals may contain compressed audio (for example, in Dolby-D/E format). If the compressed audio is synchronous with the NEXUS system, it is de-embedded or embedded as normal and is then routed over the NEXUS bus to the encoders or decoders. However, if the SDI signals are not in sync with the NEXUS, the procedure is more complex. This is because the compressed data cannot be processed by an SRC. The XHDI provides a solution for this problem; a dedicated SDI link connecting the XHDI to the Dolby encoders and decoders. After the asynchronous compressed audio has been de-embedded, the XDED decoder board decodes it asynchronously. The signals which are now available separately can then be synchronised by the SRCs on the XDED to the NEXUS bus.
The XDEE Dolby-encoder board enables the video clock frequency to be shifted to reconcile latencies. This ensures that the signals comply with the strict Dolby E phase specifications. In combination, the audio delay on the Dolby boards and the video delay provided by the XHDI also support latency compensation and phase correction.
Thanks to these features, the XHDI is the perfect solution for any application.

Signal Processor

The XHDI board features DSP for level adjustment and signal processing without relying on other boards. The input and output levels can be adjusted separately. Reverse-polarity signals can be corrected using a phase-inversion function.
If audio is to be output in 20-bit or 16-bit format rather than with the NEXUS internal 24-bit resolution, the signal quality can be significantly improved by performing the down-quantizing function with dithering enabled. In addition, user selectable noise shaping can move the noise energy up to frequency ranges less perceptible by the human ear.
Note that the signal-processing functions are not available for compressed audio.


The SMPTE 2020-1 and SMPTE 2020-2 standards support nine possible data streams per video signal. The XHDI 02 is capable of de-embedding two metadata streams and embedding two further streams at the same time. The user selects which data to work with. Metadata streams are assigned to stereo audio pairs. The user at the transmitting end makes this assignment; the receiver sees all available streams.
There are two ways of applying metadata to, or sending it from, the XHDI: as RS232 or RS422 signals using the optional XDEM interface module, or as an XTI signal. Combinations of these two modes are also supported. The major advantage of the XTI format is that metadata can be routed on the NEXUS just like audio. The XHDI exchanges them either directly with the XDED and XDEE Dolby boards or with external devices using an XTI board.
As with the audio, the metadata send channels can be delayed.
Metadata must be synchronous with the video. Since there is no synchronicity with the NEXUS system clock, there is no need to employ SRCs: The metadata-signal path can always be the same regardless of the audio-sync status.

Error Detection

To achieve maximum reliability, the input-signal status and board status is checked regularly. The control software informs the user of any issues.
Two front-panel LEDs indicate the board status (sync errors, SD/HD/3G mode, test mode, and the status of the I/O PLL).
In the case of an input-signal failure, the board switches automatically to test-generator mode. (This feature can be disabled.) The output-signal format can be set using the control software.
In case of failure of a critical system component (e.g. the XCPU board), the board enters a bypass mode automatically, ensuring that the video stream is not interrupted. The NEXUS Base Devices feature redundant power-supply units to avoid power failure.


The XHDI board is equipped with BNC or optical ports (LC, single mode).
The BNC version offers one BNC input, two parallel BNC outputs, and one BNC-through output, providing the re-clocked input signal.
Optical I/O ports are implemented as SFP modules providing LC terminals. The use of SFP modules enables the optical modules to be installed or replaced on site.
This makes adaptation for specific applications (distance, wavelength, etc.) simple.


XHDI-SFP SFP module, single-mode (1,300µm), spanning distances of up to 2 km; LC, SM, DMI

XHDI Specifications

Data formats SMPTE 259M (SD), SMPTE 292M (HD), and SMPTE 424M/425M (3G) compliant video; SMPTE 272M-AC (SD) and SMPTE 299M (HD/3G) compliant audio; SMPTE RP165 (SD) compliant EDH; SMPTE 2020-1/-2 compliant metadata
Audio data 16/20/24-bit
Audio channels 16 inputs, 16 outputs
Sample rate 48 KHz (interface), 32 to 96 KHz (system)
Metadata channels 2 inputs, 2 outputs
Metadata rate 115.2 Kbps (nominal)
Video rate 270/1,483.51/1,485 Mbps or 2,970/2,967 Mbps (nominal)
Video latency <200 pixels (no video delay set)
BNC inputs 1 BNC
Reflection attenuation: >15 dB @ 5 MHz to 1.5 GHz (typ., SMPTE 424M)
>10 dB @ 1.5 to 3 GHz (typ., SMPTE 424M)
Cable length (recommended): 100m typ., 250m max. @ 270Mbps
100m typ., 230m max. @ 1,485Mbps
100m typ., 140m max. @ 2,970Mbps
Impedance: 75 ohm
BNC outputs
2 parallel BNC ports, 1 BNC through
Reflection attenuation: >15 dB @ 5MHz to 1.5 GHz (typ., SMPTE 424M)
>10 dB @ 1.5 to 3 GHz (typ., SMPTE 424M)
Output voltage: min. 750 mVPP , typ. 800 mVPP , max. 850 mVPP
Impedance: 75 ohm
Optical input 1 LC port, single-mode, (max. rec. cable length: 10 km – example)
Optical power -20 dBm @ 9/125 µm (min.)
Overall attenuation 15 dBm (cable + connectors + splices/switches, etc., max.)
Optical output
1 LC port, single-mode, (max. rec. cable length: 10 km – example)
Center wavelength 1,310 µm (typ.)
Optical power -5 dBm @ 9/125 µm (min.)