Agile Live 7.0.0 developer reference

• 1: REST API v3
• 2: Agile Live Rendering Engine configuration documentation
• 3: Agile Live Rendering Engine command documentation
• 4: C++ SDK
• 4.1: Classes
• 4.1.1: AclLog::CommandLogFormatter
• 4.1.2: AclLog::FileLocationFormatterFlag
• 4.1.3: AclLog::ThreadNameFormatterFlag
• 4.1.4: AlignedAudioFrame
• 4.1.5: AlignedFrame
• 4.1.6: ControlDataCommon::ConnectionStatus
• 4.1.7: ControlDataCommon::Response
• 4.1.8: ControlDataCommon::StatusMessage
• 4.1.9: ControlDataSender
• 4.1.10: ControlDataSender::Settings
• 4.1.11: DeviceMemory
• 4.1.12: IControlDataReceiver
• 4.1.13: IControlDataReceiver::IncomingRequest
• 4.1.14: IControlDataReceiver::ReceiverResponse
• 4.1.15: IControlDataReceiver::Settings
• 4.1.16: IMediaStreamer
• 4.1.17: IMediaStreamer::Configuration
• 4.1.18: IMediaStreamer::Settings
• 4.1.19: IngestApplication
• 4.1.20: IngestApplication::Settings
• 4.1.21: ISystemControllerInterface
• 4.1.22: ISystemControllerInterface::Callbacks
• 4.1.23: ISystemControllerInterface::Response
• 4.1.24: MediaReceiver
• 4.1.25: MediaReceiver::NewStreamParameters
• 4.1.26: MediaReceiver::Settings
• 4.1.27: SystemControllerConnection
• 4.1.28: SystemControllerConnection::Settings
• 4.1.29: TimeCommon::TAIStatus
• 4.1.30: TimeCommon::TimeStructure
• 4.2: Files
• 4.2.1: include/AclLog.h
• 4.2.2: include/AlignedFrame.h
• 4.2.3: include/Base64.h
• 4.2.4: include/ControlDataCommon.h
• 4.2.5: include/ControlDataSender.h
• 4.2.6: include/DeviceMemory.h
• 4.2.7: include/IControlDataReceiver.h
• 4.2.8: include/IMediaStreamer.h
• 4.2.9: include/IngestApplication.h
• 4.2.10: include/IngestUtils.h
• 4.2.11: include/ISystemControllerInterface.h
• 4.2.12: include/MediaReceiver.h
• 4.2.13: include/PixelFormat.h
• 4.2.14: include/SystemControllerConnection.h
• 4.2.15: include/TimeCommon.h
• 4.2.16: include/UUIDUtils.h
• 4.3: Namespaces
• 4.3.1: ACL
• 4.3.2: AclLog
• 4.3.3: ControlDataCommon
• 4.3.4: IngestUtils
• 4.3.5: spdlog
• 4.3.6: TimeCommon
• 4.3.7: UUIDUtils
• 5: System controller config

1 - REST API v3

The following is a rendering of the OpenAPI for the Agile Live System Controller using Swagger. It does not have a backend server, so it cannot be run interactively.

This API is available from the server at the base URL /api/v3.

2 - Agile Live Rendering Engine configuration documentation

This page describes how to configure the Rendering Engine. This topic is closely related to this page on the control command protocol for the video and audio mixers.

Rendering Engine components

The Rendering Engine is an application that uses the Production Pipeline library in the base platform for transport, and adds to that media file playback, HTML rendering, a full video mixer and an audio router and audio mixer. The figure below shows a schematic of the different components and an example of how streams may transition through it.

HTML Renderers

Multiple HTML renderers can be instantiated in runtime by control panels, and in each an HTML page can be opened. Each HTML renderer produces a video stream, but no audio.

Media Players

Multiple media players can be instantiated in runtime by control panels, and in each a media file can be opened. Each media player produces a stream containing a video stream and all audio streams from the file.

Video Mixer

The Video mixer receives all video inputs into the system, i.e. streams from ingests, from HTML renderers and from media players. It outputs one or more named video output streams. The internal structure of the video mixer is defined at startup (as decribed in detail below), and controlling the video mixer is done in runtime by control panels.

Audio Router

The audio router receives all audio streams into the system, i.e. streams from ingests and from media players. It outputs a set of streams that are either mono or a stereo pair, to the audio mixer. The mappings from input streams to output streams in the router is configured in runtime by control panels.

Audio Mixer

The audio mixer takes a number of mono or stereo pair streams as inputs that we call input strips. It outputs a number of named output streams in stereo. The outputs of the audio mixer are defined at startup, and the audio mixer is controlled in runtime by control panels.

Combine outputs

The combination of video and audio outputs into full output streams from the rendering engine is configured at startup.

Configuring the Rendering Engine

Some aspects of the rendering engine can be configured statically at startup through the use of a configuration file in JSON format. Specifically, the video mixer node graph, the audio mixer outputs and the combination of video and audio outputs can be configured. As an example, here is such a JSON configuration file that we will refer to throughout this guide:

{
  "version": "1.0",
  "video": {
    "nodes": {
      "transition": {
        "type": "transition"
      },
      "chroma_key_select": {
        "type": "select"
      },
      "chroma_key": {
        "type": "chroma_key"
      },
      "chroma_key_alpha_over": {
        "type": "alpha_over"
      },
      "fade_to_black": {
        "type": "fade_to_black"
      },
      "program": {
        "type": "output"
      },
      "chroma_key_preview": {
        "type": "output"
      },
      "preview": {
        "type": "output"
      }
    },
    "links": [
      {
        "from_node": "transition",
        "from_socket": 0,
        "to_node": "chroma_key_alpha_over",
        "to_socket": 1
      },
      {
        "from_node": "transition",
        "from_socket": 1,
        "to_node": "preview",
        "to_socket": 0
      },
      {
        "from_node": "chroma_key_select",
        "from_socket": 0,
        "to_node": "chroma_key",
        "to_socket": 0
      },
      {
        "from_node": "chroma_key",
        "from_socket": 0,
        "to_node": "chroma_key_alpha_over",
        "to_socket": 0
      },
      {
        "from_node": "chroma_key_alpha_over",
        "from_socket": 0,
        "to_node": "fade_to_black",
        "to_socket": 0
      },
      {
        "from_node": "fade_to_black",
        "from_socket": 0,
        "to_node": "program",
        "to_socket": 0
      },
      {
        "from_node": "chroma_key",
        "from_socket": 0,
        "to_node": "chroma_key_preview",
        "to_socket": 0
      }
    ]
  },
  "audio": {
    "outputs": [
      {
        "name": "main",
        "channels": 2
      },
      {
        "name": "aux1",
        "channels": 2,
        "follows": "main"
      },
      {
        "name": "aux2",
        "channels": 2
      }
    ]
  },
  "output_mapping": [
    {
      "name": "program",
      "video_output": "program",
      "audio_output": "main",
      "feedback_input_slot": 1001,
      "stream": true
    },
    {
      "name": "preview",
      "video_output": "preview",
      "audio_output": "",
      "feedback_input_slot": 1002,
      "stream": false
    },
    {
      "name": "aux1",
      "video_output": "program",
      "audio_output": "aux1",
      "feedback_input_slot": 0,
      "stream": true
    },
    {
      "name": "chroma_key_preview",
      "video_output": "chroma_key_preview",
      "audio_output": "aux2",
      "feedback_input_slot": 100,
      "stream": true
    }
  ]
}

Video Mixer node graph

The video mixer is defined as a tree graph of nodes that work together in sequence to produce one or several video outputs. Each node performs a specific action, and has zero or more input sockets and zero or more output sockets. Links connect output sockets on one node to input sockets on other nodes. Each node is named and the name is used to control the node in runtime. The node tree configuration is specified in the video section of the JSON file, which contains two parts: the list of nodes and the list of links connecting the nodes.

The following is a graphical representation of the video mixer node graph configuration in the JSON example file above, with two input nodes, three processing nodes and three output nodes, and links in between:

Video nodes block

The nodes object of the video block is a JSON object where the names/keys in the object is the unique name of the node. The name is used to refer to the node when operating the mixer later on, so names that are easy to understand what they are supposed to be used for in the production is recommended. The name can only contain lower case letters, digits, _ and -. Space or other special characters are not allowed.

Each node is an JSON object with parameters defining the node properties. Here the parameter type defines which type of node it is. The supported types are listed below. They can be divided into three groups depending on if they provide input to the graph, output from the graph or is a processing node, placed in the middle of the graph.

Input nodes

The input nodes take their input from the input slots of the Rendering Engine, which contains the latest frame for all connected sources (this includes connected cameras, HTML graphics, media players etc.) The nodes in this group does not have any input sockets, as they take the input frames from the input slots of the Rendering Engine. Which slot to take the frames from is dynamically controlled during runtime.

Alpha combine node (alpha_combine)
Input sockets: 0 (Sources are taken from the input slots of the mixer)
Output sockets: 1
The alpha combine node takes two inputs and combines the color from one of them with the alpha from the other one. The node features multiple modes that can be set during runtime to either pick the alpha channel of the second video input, or to take any of the other R, G and B channels, or to average the RGB channels and use as alpha for the output. This node is useful in case videos with alpha is provided from SDI sources, where the alpha must be sent as a separate video feed and then combined with the fill/color source in the mixer.

Select node (select)
Input sockets: 0 (Source is taken from the input slots of the mixer)
Output sockets: 1
The select node simply selects an input slot from the Rendering Engine and forwards that video stream. Which input slot to forward is set during runtime. The node is a variant of the transition node, but does not support the transitions supported by the transition node.

Transition node (transition)
Input sockets: 0 (Sources are taken from the input slots of the mixer)
Output sockets: 2 (One with the program output and one with the preview output)
The transition node takes two video streams from the Rendering Engine’s input slots, one to use as the program and one as the preview output stream. These are output through the two output sockets. During runtime this node can be used to make transitions such as wipes and fades between the selected program and preview.

Output nodes

This group only contains a single node, used to mark a point where processed video can exit the graph and be used in output mappings.

Output node (output)
Input sockets: 1
Output sockets: 0 (Output is sent out of the video mixer)
The output node marks an output point in the video mixer’s graph, where the video stream can be used by the output mapping to be included in the Rendering Engine’s Output streams, or as streams to view in the multi-viewer. The node takes a single input and makes it possible to use that video feed outside the video mixer. Output nodes can be used both to output the program and preview feeds of a video mixer, but also to mark auxiliary outputs, as in the example above, where chroma_key_preview is output to be included in the graph to be able to view the result of the chroma keying, without the effect being keyed on to the program output.

Processing nodes

The processing nodes take their input from another node’s output and outputs a result that is sent to another node’s input. They are therefore placed in the middle of the graph, after nodes from the input group and before output nodes.

Alpha over node (alpha_over)
Input sockets: 2 (Index 0 for overlay video and index 1 for background video)
Output sockets: 1
The alpha over node composites the overlay video input on top of the background video input. The alpha of the overlay video input is taken into consideration. During runtime, this node can be controlled to show or not to show the overlay, and to fade the overlay in or out. This node is useful to composite things such as graphics or chroma keyed video onto a background video.

Chroma key node (chroma_key)
Input sockets: 1
Output sockets: 1
The chroma key node takes an input video stream and performs chroma keying on it based on parameters set during runtime. The video output will have the alpha channel (and in some cases also the color channels) altered. The result of this node can then be composited on top of a background using an Alpha over node.

Fade to black node (fade_to_black)
Input sockets: 1
Output sockets: 1
The fade to black node takes a single input video stream and can fade that video stream to and from black. This node is normally used as the last node before the main program output node, to be able to fade to and from black at the beginning and end of the broadcast.

Transform node (transform)
Input sockets: 1
Output sockets: 1
The transform node takes an input video stream and transform the result inside the visible canvas. This node can be configured during runtime to scale and move the input video. This node is useful for picture-in-picture effects, or to move a chroma key node output to the lower corner of the frame.

Video delay node (video_delay)
Input sockets: 1
Output sockets: 1
The video delay node is used to delay the video by a given number of frames. The number of frames to delay is controlled during runtime. This node is useful whenever the need for dynamically delay a video stream arises, for example in case an external audio mixer is used, which comes with a delay of some frames.

Video links block

The links array in the video block is a list of links between video nodes. The video frames are fed in one direction from node to node via these links. An input socket on a node can only have one connected link. The output sockets on a node can have multiple connected links.

Each block contains the following keys:

• from_node - The name of the node in the nodes object, from which the link is receiving frames from
• from_socket - The index of the output socket in the node from which this link originates
• to_node - The name of the node in the nodes object, to which the link is sending frames to
• to_socket - The index of the input socket in the node to which this link connects

Some examples

The simplest video mixer node graph imaginable would be a select node feeding an output node. This mixer would only be able to select one of the inputs and output it unaltered, like a video router would do:



The JSON file section for this is:

"video": {
  "nodes": {
    "input_select": {
      "type": "select"
    },
    "program": {
      "type": "output"
    }
  },
  "links": [
    {
      "from_node": "input_select",
      "from_socket": 0,
      "to_node": "program",
      "to_socket": 0
    }
  ]
}

A slightly more advanced node graph would be to use a transition node and two outputs, one for the program out and one for the preview:



The JSON file section for this is:

"video": {
  "nodes": {
    "transition": {
      "type": "transition"
    },
    "program": {
      "type": "output"
    },
    "preview": {
      "type": "output"
    }
  },
  "links": [
      {
        "from_node": "transition",
        "from_socket": 0,
        "to_node": "program",
        "to_socket": 0
      },
      {
        "from_node": "transition",
        "from_socket": 1,
        "to_node": "preview",
        "to_socket": 0
      },
  ]
}

Audio outputs

The audio block of the configuration file defines the properties of the audio mixer. The JSON object contains a list called outputs which lists the outputs of the audio mixer and the configuration of each output. Each output has these parameters:

• name - The unique name of this audio output, used to refer to if from the output_mapping block
• channels - The number of audio channels of this output
• follows - (Optional parameter) Used to identify the audio output this output is a post fader aux output for. If the output does not have this parameter set, it is considered a main output

The follows parameter is used to set the output in “post fader aux send” mode. This is used to create extra aux outputs from the audio mixer which are used for mix minus, i.e. where you want the program output, but with some specific audio source(s) removed. When follows is set, the volume of that output will follow the volume of the main bus and scale that volume. If the volume fader of the aux bus is up, it will send the same volume of audio of that strip as the main bus does. So when the main bus volume is turned down, the same thing will happen automatically in the aux bus. If an aux bus’ fader is down on the other hand, that strip will not contribute to the aux bus at all. The aux bus volume faders can therefore be used to remove audio (or scale the volume up or down), but cannot be used to add more audio strips compared to the main bus it is following.

The input routing and configuration of the strips is made via the control command API.

Combine outputs

The output_mapping block is used to define outputs of the entire Rendering Engine, by combining outputs from the video and audio mixers and configuring where those should be sent. The output mapping is an array of output mappings. Each mapping has the following parameters:

• name - The unique name of the output mapping. This will be displayed in the REST API both for sources being streamed and sources with feedback streams that can be included in the multi-view
• video_output - The name of the video mixer node of type output to get the video stream from. Leave empty, or omit the key to create an audio-only output
• audio_output - The name of the audio output to get the audio stream from. Leave empty, or omit the key to create a video-only output
• feedback_input_slot - The input slot to use to feed this output back to the multi-view. The REST API may then refer to this input slot to include this output in a multi-view view. Value must be >= 100 as input slots up to 99 are reserved for “regular” sources. Use 0 to disable feedback of this stream.
• stream - Boolean value to tell if the output should be streamable and visible as an output in the REST API. If set to false the output will not turn up as an Pipeline Output in the REST API.

The following is a graphical visualisation of the output mappings in the JSON example configuration file above:

3 - Agile Live Rendering Engine command documentation

This page describes the commands for controlling the video mixer, audio router, audio mixer, HTML renderers and media players in the Agile Live Rendering Engine. This topic is closely related to this page on how to configure the rendering engine at startup.

Command protocol

All commands to the Agile Live Rendering Engine are sent as human readable strings and are listed below. The Rendering Engine expect no line termination, meaning that you do not need to append any type of new line (\n) characters, or string termination (\0) characters.

Each subsystem of the rendering engine has their own set of commands, prefixed by the subsystem name. As can be seen in the image below, the prefixes of the different subsystems are:

• html for the html renderer instances
• media for the media playback instances
• video for the video mixer
• audio map for the audio router
• audio for the audio mixer (if using the built-in mixer)

Video mixer commands

Background

The video mixer is built as a tree graph of processing nodes, please see this page for further information on the video mixer node graph. The names of the nodes defined in the node graph are used as the first word in the control commands as the way to address the specific node. On this page we will specify the detailed protocol for each node type.

There are eight different node types.

• The Transition node is used to pick which input slots to use for the program and preview output. The node also supports making transitions between the program and preview.
• The Select node simply forwards one of the available sources to its output.
• The Alpha combine node is used to pick two input slots (or the same input slot) to copy the color from one input and combine it with some information from the other input to form the alpha. The color will just be copied from the color input frame, but there are several modes that can be used to produce the alpha channel in the output frame in different ways. This is known as “key & fill” in broadcasting.
• The Alpha over node is used to perform an “Alpha Over” operation, that is to put the overlay video stream on top of the background video stream, and let the background be seen through the overlay depending on the alpha of the overlay. The node also features fading the graphics in and out by multiplying the alpha channel by a constant factor.
• The Transform node takes one input stream and transforms it (scaling and translating) to one output stream.
• The Chroma Key node takes one input stream, and by setting appropriate parameters for the keying, it will remove areas with the key color from the incoming video stream both affecting the alpha and color channels
• The Fade to black node takes one input stream, which it can fade to or from black gradually, and then outputs that stream.
• The Output node has one input stream and will output that stream out from the video mixer, back to the Rendering Engine. It has no control commands

To reset the runtime configuration of all video nodes to their default state, the following command is used:

• video reset - Reset the runtime configuration of all video mixer nodes

Transition

The transition type of node can be controlled with the following commands:

• video <node_name> cut <input_slot> - Cut (hot punch) the program output to the given input slot. This won’t affect the preview output.
• video <node_name> preview <input_slot> - Preview the given input slot. This won’t affect the program output.
• video <node_name> cut - Cut between the current program and preview outputs, effectively swapping their place.
• video <node_name> panic <input_slot> - Cut (hot punch) both the preview and program output to the given input slot instantly for all rendering engines that are receiving this command, disregarding any timing differences!

Some transition commands last over a duration of time, for example wipes. These can be performed either automatically or manually. The automatic mode works by the operator first selecting the type of transition, for instance a fade, setting the preview to the input slot to fade to and then trigger the transition at the right time with a set duration for the transition. In manual mode the exact “position” of the transition is set by the control panel. This is used for implementing T-bars, where the T-bar repeatedly sends the current position of the bar. In the manual mode, the transition type is set before the transition begins, just as in the automatic mode. Note that an automatic transition will be overridden in case the transition “position” is manually set, by interrupting the automatic transition and jumping to the manually set position.

• video <node_name> type <type> - Set the transition type to use for transitions hereafter. Valid modes are:
  • fade - Fade (mix) from one source to the next
  • wipe_left - Wipe from one source to the next, wipe going from right side of the screen to the left
  • wipe_right - Wipe from one source to the next, wipe going from left side of the screen to the right
• video <node_name> auto <duration_ms> - Perform an automatic transition from the current program output to the current preview output, lasting for <duration_ms> milliseconds. The currently set transition type is used.
• video <node_name> factor <factor> - Manually set the position/factor of the transition. <factor> should be between 0.0 and 1.0, where 0.0 is the value before the transition starts and 1.0 is the end of the transition. Note that setting this value to 1.0 will effectively swap place of program and preview output, resetting the transition factor to 0.0 internally. Control panels using this command must take this into consideration to not cause a sudden jump in the transition once the physical control is moved again. Sending this command will interrupt any ongoing automatic transition.

Select

The select type of node only has one control command:

• video <node_name> cut <input_slot> - Choose which input slot to forward. Changing looks like a cut.

Alpha Combine

The alpha combine node type can be controlled with these commands:

• video <node_name> color <input_slot> - Set which input slot to copy the color data from in the Alpha combine node
• video <node_name> alpha <input_slot> - Set which input slot to use to create the alpha channel of the output frame in the Alpha combine node
• video <node_name> mode <mode> - Set which way the alpha channel of the output frame will be produced. Valid modes are:
  • copy-r - Copy the R-channel of the input “alpha frame” and use it as alpha for the output
  • copy-g - Copy the G-channel of the input “alpha frame” and use it as alpha for the output
  • copy-b - Copy the B-channel of the input “alpha frame” and use it as alpha for the output
  • copy-a - Copy the A-channel of the input “alpha frame” and use it as alpha for the output
  • average-rgb - Take the average of the R, G and B channels of the input “alpha frame” and use it as alpha channel for the output

Alpha over

The alpha over node type can be controlled with these commands:

• video <node_name> factor <factor> - Manually set the transparency multiplier for the Alpha over node. The value <factor> should be a float between 0.0 and 1.0, where 0.0 means that the overlay is completely invisible and 1.0 means that the overlay is fully visible (where it should be visible according to the overlay image’s alpha channel). Default is 0.0.
• video <node_name> fade to <duration_ms> - Start an automatic fade from the current factor to a fully visible overlay over the duration of <duration_ms> milliseconds.
• video <node_name> fade from <duration_ms> - Start an automatic fade from the current factor to a fully invisible overlay over the duration of <duration_ms> milliseconds.

Transform

The transform node type can be controlled with these commands:

• video <node_name> scale <factor> - Set the scale of the output size, relative the original size. Parameter factor must be > 0. Default is 1.0.
• video <node_name> x <position> - Set the X offset of the top left corner of the output, relative the top left corner of the frame. Position is measured in percent of frame width, where 0.0 is the left side of the frame and 1.0 is the right side of the frame. Values outside of the range 0, 1 are valid. Default is 0.0.
• video <node_name> y <position> - Set the Y offset of the top left corner of the output, relative the top left corner of the frame. Position is measured in percent of frame height, where 0.0 is the top of the frame and 1.0 is the bottom of the frame. Values outside of the range 0, 1 are valid. Default is 0.0.

Chroma Key

The chroma key node type can be controlled with these commands:

• video <node_name> r <value> - Set the red component of the chroma key color. Should be a float between 0.0 and 1.0.
• video <node_name> g <value> - Set the green component of the chroma key color. Should be a float between 0.0 and 1.0.
• video <node_name> b <value> - Set the blue component of the chroma key color. Should be a float between 0.0 and 1.0.
• video <node_name> distance <value> - Set the distance parameter, ranging from 0.0 to 1.0, telling how far from the key color is also considered a part of the key. Larger values will result in more colors being included, 0.0 means only the exact key color is considered.
• video <node_name> falloff <value> - Set the falloff parameter, ranging from 0.0 to 1.0, which makes edges smoother. A lower value means harder edges between the key color and the parts to keep, a higher value means a more smooth falloff between removed and kept parts.
• video <node_name> spill <value> - Set the color spill reduction parameter, ranging from 0.0 to 1.0, which will desaturate the key color in the parts of the image that are left. Useful for hiding green rims around objects in from of a green screen etc. 0.0 means off and a higher value means more colors further away from the chroma key color will be desaturated.
• video <node_name> color_picker <x_factor> <y_factor> <square_size> - Set the chroma key by sampling the pixels covered by the square area of size square_size x square_size. The position x,y is set by <x_factor> * width of video and <y_factor> * height of video. <x_factor> & <y_factor> should be a float between 0.0 and 1.0. The <square_size> should be an integer > 0.0.
• video <node_name> alpha_as_video <value> - Instead of writing an alpha mask, set the alpha value as the RGB value, creating a grayscale image of the mask. Value should be true or false.
• video <node_name> show_key_color <value> - Draw a 100x100 box in upper left corner with the current chroma key (RGB) value. Value should be true or false.
• video <node_name> show_color_picker <value> - Draw a pink border around the area where the color picker will sample a new chroma key. Value should be true or false.

Fade to black

The fade to black node can be controlled via the following commands:

• video <node_name> factor <factor> - Manually set the blackness factor on a scale from 0.0 to 1.0, where 0.0 means not black and 1.0 means fully black output.
• video <node_name> fade to <duration_ms> - Start an automatic fade to a fully black output over the duration of <duration_ms> milliseconds.
• video <node_name> fade from <duration_ms> - Start an automatic fade to a fully visible (non-black) output over the duration of <duration_ms> milliseconds.

Audio Router commands

The audio router is used to route specific audio channels from an input slot of the Rendering Engine to an input channel strip on the audio mixer. By default, it has no audio routes set up. The router can be controlled with the following commands:

• audio map <input_slot> <channel> mono <input_strip> - Map the <channel> from input slot <input_slot> as a mono channel into input strip <input_strip> of the audio mixer
• audio map <input_slot> <left_channel> stereo <input_strip> - Map the <left_channel> and <left_channel> + 1 from input slot <input_slot> as a stereo pair into input strip <input_strip> of the audio mixer
• audio map reset <input_strip> - Remove the current mapping from <input_strip> of the audio mixer.

To reset both the audio mixer and the audio router to their default state, including removal of all configured input strips, the following command can be used:

• audio reset - Reset both the audio mixer and audio router

Audio mixer commands

The audio mixer and audio router have the following data path:

In the example picture above, three sources are connected to the input slots and four mono or stereo audio streams are picked as input to the audio mixer. The audio mixer is completely separate from the video mixer, so switching image in the video mixer will not change the audio mixer in any way. The internal audio mixer can be controlled using the following commands:

Pre-gain

To control the pre-gain of each input strip use:

• audio gain <input_strip> <level_dB> - Set the pre-gain level of a specific input in dB. Parameter <level> should be a floating point number. Default is 0.0.

Low-pass filter

To control the low-pass filter of each input strip use:

• audio lpf <input_strip> freq <value> - Set the cutoff frequency for the low-pass filter in an input strip. <value> is the frequency in Hz as a floating point number. The frequency range is 20 Hz to 20 kHz. Setting a value of 20 kHz or higher will bypass the filter.
• audio lpf <input_strip> q <value> - Set the q-value for the low-pass filter in an input strip. The q-value controls the steepness of cutoff curve with a higher value being a steeper slope. The value is a floating point number ranging from 0.4 to 6.0. The default value is 0.7.

High-pass filter

To control the high-pass filter of each input strip use:

• audio hpf <input_strip> freq <value> - Set the cutoff frequency for the high-pass filter in an input strip. <value> is the frequency in Hz as a floating point number. The frequency range is 20 Hz to 20 kHz. Setting a value of 20 Hz or lower will bypass the filter.
• audio hpf <input_strip> q <value> - Set the q-value for the high-pass filter in an input strip. The q-value controls the steepness of cutoff curve with a higher value being a steeper slope. The value is a floating point number ranging from 0.4 to 6.0. The default value is 0.7.

Parametric equalizer

The parametric equalizer has ten bands, indexed from 0 to 9. By default, all bands are disabled (type “none”) with a gain of 0 dB, a Q value of 0.707, and a frequency of 1 kHz. To control the parametric equalizer of each input strip use:

• audio eq <input_strip> <band> freq <value> - Set the frequency of a specific band in the parametric equalizer. Here <band> is the zero-based index of the band. <value> is a floating point number with the frequency in Hz. For peak, notch, and band-pass filters this is the center frequency. For low-pass, high-pass, low-shelf, and high-shelf filters this is the corner frequency.
• audio eq <input_strip> <band> gain <value> - Set the gain of a specific band in the parametric equalizer. Here <band> is the zero-based index of the band. <value> is a floating point number with the gain in dB. The gain parameter only has effect on peaking and shelving filters.
• audio eq <input_strip> <band> q <value> - Set the q-value of a specific band in the parametric equalizer to shape the falloff of the band. Here <band> is the zero-based index of the band. <value> is a floating point number with the q-value where a higher value means a more pointy curve. The default value is 0.707.
• audio eq <input_strip> <band> type <type> - Set the filter type to use for a specific band in the parametric equalizer. Here <band> is the zero-based index of the band. The available values for <type> are:
  • none: Bypass audio without any changes
  • lowpass: Low-pass filter at the current frequency. Gain has no effect.
  • highpass: High-pass filter at the current frequency. Gain has no effect.
  • bandpass: Band-pass filter at the current frequency. Gain has no effect.
  • lowshelf: Low-shelf filter. Audio frequencies below the currently set value are modified by the current gain value.
  • highshelf: High-shelf filter. Audio frequencies above the currently set value are modified by the current gain value.
  • peak: Peak filter. Frequencies around the currently set value are modified by the current gain value.
  • notch: Notch filter. Frequencies around the currently set value are reduced greatly. Gain has no effect.

Panning

To control the panning of each input strip use:

• audio pan <input_strip> <factor> - Pan the mono or stereo input of an input strip to the left or right. The range of <factor> is between -1.0 and +1.0, where 0.0 means center (no panning), negative values means more to the left and positive value more to the right. Values outside this range will be clamped. A value of +/- 1.0 means that there will only be audio in the left/right channel.

Dynamic range compressor

To control the dynamic range compressor of each input strip use:

• audio comp <input_strip> threshold <value> - Set the threshold for activation of the compressor. The threshold is a negative dB value ranging from -30 dB to 0 dB. The volume of audio which is above the threshold value will be reduced (compressed). The default value is 0 dB, i.e. only compression if the audio signal is overloaded.
• audio comp <input_strip> ratio <value> - Set the compression ratio for audio exceeding the loudness threshold. The value is the numerator in the compression ratio <value>:1. For instance, if the value is set to 4, the compression ratio is 4:1 and volume overshoot above the threshold will be scaled down to 25 %. The ratio numerator is a floating point number in the range from 1.0 to 24.0, with 4.0 as the default value.
• audio comp <input_strip> knee <value> - Set the width of the soft knee in decibels. Instead of simply turning the compression completely on or off at the threshold, the knee defines a volume range in which the compression ratio follows a curve, the “knee”. The knee is a floating point number between 0 dB and 30 dB, with a default value of 2.5 dB.
• audio comp <input_strip> attack <value> - Set the attack time of the compressor in milliseconds. The attack time determines how long it takes to reach the full compression after the threshold has been exceeded. The value is a floating point number in the range from 0.1 ms to 120 ms. The default value is 50 ms.
• audio comp <input_strip> release <value> - Set the release time of the compressor in milliseconds. The release time determines how long it takes to return to zero compression when the volume is below the compression threshold. The value is a floating point number in the range from 10 ms to 1000 ms. The default value is 200 ms.
• audio comp <input_strip> gain <value> - Set the make-up gain i decibels. Since the compression filter lowers the volume of louder audio sections it can be desirable to increase the gain after the filtering. The gain value increases the audio volume with the specified number of decibels. The value is a floating point number in the range from 0 dB to 24 dB. The default value is 0 dB.

Volume and master volume

To control the volume faders and master volume faders use:

• audio volume <input_strip> <output> <level> - Set the volume level of a specific input in proportion to its original strength for a given output. Parameter <level> should be a floating point number, where 0.0 means that the input slot is not mixed into the output at all and 1.0 means that the volume of the input slot is kept as is for the output. The signal can also be amplified by using values > 1.0. Default is 0.0.
• audio mastervolume <output> <level> - Set the master volume level of an output, before the audio is clamped and converted from floating point samples to 16-bit integer samples. This can be used to avoid clipping of the audio when mixing multiple input sources. Parameter <level> is a floating point number where 0.0 means that all audio output is off, 1.0 means that the volume is not changed and any value above 1.0 means that the master volume is amplified. Default is 1.0.
• audio volume <input_strip> <output> auto <level> <time_ms> - Do automatic transition of volume from the current volume to the requested level over time_ms milliseconds. The volume will be linearly faded over time in the dB scale.
• audio mastervolume <output> auto <level> <time_ms> - Do automatic transition of the master volume from the current volume to the requested level over time_ms milliseconds. The volume will be linearly faded over time in the dB scale.
• audio volume <input_strip> <output> mute <value> - Mute or unmute a specific input strip for a specific output bus. This operation will not change any gain or volume fader of the strip. I.e. unmuting will restore the same audio level as before mute was sent. Value should be a boolean, true or false. Use true to mute and false to unmute.
• audio mastervolume <output> mute <value> - Mute or unmute all audio in a specific output bus. This operation will not change any gain or volume fader of any strip. I.e. unmuting will restore the same audio levels as before mute was sent. Value should be a boolean, true or false. Use true to mute and false to unmute.

Reset all audio filters

To reset both the audio mixer and the audio router to their default state, including removal of all configured input strips, the following command can be used:

• audio reset - Reset both the audio mixer and audio router

HTML rendering

The Rendering Engine also features a built-in HTML renderer which uses the Chromium web browser engine to render HTML pages. The following commands can be used to create and control the HTML browser instances:

• html create <input_slot> <width> <height> - Create a new HTML browser instance with the canvas size of width x height pixels and output the rendered frames to the given input slot
• html close <input_slot> - Close the HTML browser instance on the given input slot
• html load <input_slot> <url> - Load a new URL in the HTML browser on the given input slot
• html execute <input_slot> <java script> - Execute JavaScript in the HTML browser on the given input slot. The JavaScript snippet might span over multiple lines and may contain spaces.
• html reset - Close all open HTML browsers

Media player

The Rendering Engine can create media player instances to play video and audio files from the hard drive of the machine running the Rendering Engine. It is up to the user of the API to ensure the files are uploaded to the machines running the Rendering Engine(s) before trying to run them. The media players use the FFmpeg library to demux and decode the media files, so most files supported by FFmpeg should work. The media players each have three state parameters: start, duration and looping, to mark a section of the video to loop, or a last frame to stop at in case looping is set to false. The following commands can be used to create and control the media player instances:

• media create <input_slot> <filename> - Create a new media player instance with the given media file and output the rendered frames to the given input slot.
• media load <input_slot> <filename> - Load another file into an existing media player on the given input slot. This will reset the start, duration and looping parameters of that media player.
• media close <input_slot> - Close the media player instance on the given input slot
• media play <input_slot> - Start/resume playing the currently loaded media file in the media player on the given input slot
• media pause <input_slot> - Pause the media player on the given input slot
• media seek <input_slot> <time_ms> - Seek the media player on the given input slot to a time point time_ms milliseconds from the start of the media file and pause the playback at the first frame. Use media play <input_slot> to resume the playback
• media set_start <input_slot> <start_time_ms> - Set the start time of the looping section of the media player on the given input slot to a time point start_time_ms milliseconds from the start of the media file. It will still be possible to seek before this time point and play the video from there.
• media set_duration <input_slot> <duration_ms> - Set the duration of the playing/looping section of the media player on the given input slot to duration_ms milliseconds from the start_time_ms of the media file. The duration will be clamped in case it spans past the end of the media file. A running media player will either stop at this point, or loop the video depending on the state of the looping parameter
• media set_looping <input_slot> <true/false> - Set if the media should loop from the start time when the media player on the given input slot hits either the end of the looping section defined by the start time and duration, or the end of the media file. If set to false, the media player will stop at the end of the section or at the end of the file instead of looping it from the start point.
• media reset - Close all open media players

4 - C++ SDK

4.1 - Classes

Classes

• namespace ACL
• namespace AclLog
  A namespace for logging utilities.
  • class CommandLogFormatter
    This class is used to format log entries for Rendering Engine commands.
  • class FileLocationFormatterFlag
    A custom flag formatter which logs the source file location between a par of “[]”, in case the location is provided with the log call.
  • class ThreadNameFormatterFlag
• struct AlignedAudioFrame
  AlignedAudioFrame is a frame of interleaved floating point audio samples with a given number of channels.
• struct AlignedFrame
  A frame of aligned data that is passed to the rendering engine from the MediaReceiver. A DataFrame contains a time stamped frame of media, which might be video, audio and auxiliary data such as subtitles. A single DataFrame can contain one or multiple types of media. Which media types are included can be probed by nullptr-checking/size checking the data members. The struct has ownership of all data pointers included. The struct includes all logic for freeing the resources held by this struct and the user should therefore just make sure the struct itself is deallocated to ensure all resources are freed.
• namespace ControlDataCommon
  • struct ConnectionStatus
    Connection status struct containing information about a connection event.
  • struct Response
    A response from a ControlDataReceiver to a request. The UUID tells which receiver the response is sent from.
  • struct StatusMessage
    A status message from a ControlDataReceiver. The UUID tells which receiver the message is sent from.
• class ControlDataSender
  A ControlDataSender can send control signals to one or more receivers using a network connection. A single ControlDataSender can connect to multiple receivers, all identified by a UUID. The class is controlled using an ISystemControllerInterface; this interface is responsible for setting up connections to receivers. The ControlDataSender can send asynchronous requests to (all) the receivers and get a response back. Each response is identified with a request ID as well as the UUID of the responding receiver. The ControlDataSender can also receive status messages from the receivers.
  • struct Settings
    Settings for a ControlDataSender.
• class DeviceMemory
  RAII class for a CUDA memory buffer.
• class IControlDataReceiver
  IControlDataReceiver is the interface class for the control data receiver. An IControlDataReceiver can receive messages from a sender or other IControlDataReceivers using a network connection. It can also connect to and forward the incoming request messages to other receivers. The connections to the sender and the other receivers are controlled by an ISystemControllerInterface instance. The ControlDataReceiver has a receiving or listening side, as well as a sending side. The listening side can listen to one single network port and have multiple ControlDataSenders and ControlDataReceivers connected to that port to receive requests from them. On the sending side of the ControlDataReceiver, it can be connected to the listening side of other ControlDataReceivers, used to forward all incoming messages to that receiver, as well as sending its own requests.
  • struct IncomingRequest
    An incoming request to this ControlDataReceiver.
  • struct ReceiverResponse
    A response message to a request.
  • struct Settings
    Settings for a ControlDataReceiver.
• class IMediaStreamer
  IMediaStreamer is an interface class for MediaStreamers, that can take a single stream of uncompressed video and/or audio frames and encode and output it in some way. This output can either be a stream to a network or writing down the data to a file on the hard drive. This class is configured from two interfaces. The input configuration (input video resolution, frame rate, pixel format, number of audio channels…) is made through this C++ API. The output stream is then started from the System Controller. Any of these configurations can be made first. The actual stream to output will start once the first call to.
  • struct Configuration
    The input configuration of the frames that will be sent to this MediaStreamer. The output stream configuration is made from the System controller via the ISystemControllerInterface.
  • struct Settings
    Settings used when creating a new MediaStreamer.
• class ISystemControllerInterface
  An ISystemControllerInterface is the interface between a component and the System controller controlling the component. The interface allows for two-way communication between the component and the system controller by means of sending requests and getting responses. Classes deriving from the ISystemControllerInterface should provide the component side implementation of the communication with the system controller. This interface can be inherited and implemented by developers to connect to custom system controllers, or to directly control a component programmatically, without the need for connecting to a remote server.
  • struct Callbacks
    A struct containing the callbacks that needs to be registered by the component using this interface.
  • struct Response
    A response to a request, consists of a status code and an (optional) parameters JSON object.
• class IngestApplication
  • struct Settings
• namespace IngestUtils
• class MediaReceiver
  A MediaReceiver contains the logic for receiving, decoding and aligning incoming media sources from the Ingests. The aligned data is then delivered to the Rendering Engine which is also responsible for setting up the MediaReceiver. The MediaReceiver has a builtin multi view generator, which can create output streams containing composited subsets of the incoming video sources. This class is controlled using an ISystemControllerInterface provided when starting it.
  • struct NewStreamParameters
    A struct containing information on the format of an incoming stream.
  • struct Settings
    Settings for a MediaReceiver.
• class SystemControllerConnection
  An implementation of the ISystemControllerInterface for a System controller residing in a remote server. The connection to the server uses a Websocket.
  • struct Settings
    Settings for a SystemControllerConnection.
• namespace TimeCommon
  • struct TAIStatus
  • struct TimeStructure
• namespace UUIDUtils
  A namespace for UUID utility functions.
• namespace spdlog

Updated on 2024-04-08 at 15:15:27 +0200

4.1.1 - AclLog::CommandLogFormatter

AclLog::CommandLogFormatter

This class is used to format log entries for Rendering Engine commands.

#include <AclLog.h>

Inherits from spdlog::formatter

Public Functions

Public Functions Documentation

function CommandLogFormatter

CommandLogFormatter()

function format

void format(
    const spdlog::details::log_msg & msg,
    spdlog::memory_buf_t & dest
) override

function clone

std::unique_ptr< spdlog::formatter > clone() const override

Updated on 2024-04-08 at 15:15:27 +0200

4.1.2 - AclLog::FileLocationFormatterFlag

AclLog::FileLocationFormatterFlag

A custom flag formatter which logs the source file location between a par of “[]”, in case the location is provided with the log call.

#include <AclLog.h>

Inherits from spdlog::custom_flag_formatter

Public Functions

Public Functions Documentation

function format

inline void format(
    const spdlog::details::log_msg & msg,
    const std::tm & ,
    spdlog::memory_buf_t & dest
) override

function clone

inline std::unique_ptr< custom_flag_formatter > clone() const override

Updated on 2024-04-08 at 15:15:27 +0200

4.1.3 - AclLog::ThreadNameFormatterFlag

AclLog::ThreadNameFormatterFlag

Inherits from spdlog::custom_flag_formatter

Public Functions

Public Functions Documentation

function format

inline void format(
    const spdlog::details::log_msg & ,
    const std::tm & ,
    spdlog::memory_buf_t & dest
) override

function clone

inline std::unique_ptr< custom_flag_formatter > clone() const override

Updated on 2024-04-08 at 15:15:27 +0200

4.1.4 - AlignedAudioFrame

AlignedAudioFrame

AlignedAudioFrame is a frame of interleaved floating point audio samples with a given number of channels.

#include <AlignedFrame.h>

Public Attributes

Public Attributes Documentation

variable mSamples

std::vector< float > mSamples;

variable mNumberOfChannels

uint8_t mNumberOfChannels = 0;

variable mNumberOfSamplesPerChannel

uint32_t mNumberOfSamplesPerChannel = 0;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.5 - AlignedFrame

AlignedFrame

A frame of aligned data that is passed to the rendering engine from the MediaReceiver. A DataFrame contains a time stamped frame of media, which might be video, audio and auxiliary data such as subtitles. A single DataFrame can contain one or multiple types of media. Which media types are included can be probed by nullptr-checking/size checking the data members. The struct has ownership of all data pointers included. The struct includes all logic for freeing the resources held by this struct and the user should therefore just make sure the struct itself is deallocated to ensure all resources are freed.

#include <AlignedFrame.h>

Public Functions

Public Attributes

Public Functions Documentation

function AlignedFrame

AlignedFrame() =default

function ~AlignedFrame

~AlignedFrame() =default

function AlignedFrame

AlignedFrame(
    AlignedFrame const & 
) =delete

function operator=

AlignedFrame & operator=(
    AlignedFrame const & 
) =delete

function makeShallowCopy

std::shared_ptr< AlignedFrame > makeShallowCopy() const

Make a shallow copy of this AlignedFrame (video and audio pointers will point to the same video and audio memory buffers as in the frame copied from)

Return: A pointer to a new frame, which is a shallow copy of the old one, pointing to the same video and audio memory buffers

Public Attributes Documentation

variable mCaptureTimestamp

int64_t mCaptureTimestamp = 0;

The TAI timestamp in microseconds since the TAI epoch when this frame was captured by the ingest

variable mRenderingTimestamp

int64_t mRenderingTimestamp = 0;

The TAI timestamp in microseconds since the TAI epoch when this frame should be delivered to the rendering engine

variable mVideoFrame

std::shared_ptr< DeviceMemory > mVideoFrame = nullptr;

variable mPixelFormat

PixelFormat mPixelFormat = PixelFormat::kUnknown;

variable mFrameRateN

uint32_t mFrameRateN = 0;

variable mFrameRateD

uint32_t mFrameRateD = 0;

variable mWidth

uint32_t mWidth = 0;

variable mHeight

uint32_t mHeight = 0;

variable mAudioFrame

AlignedAudioFramePtr mAudioFrame = nullptr;

variable mAudioSamplingFrequency

uint32_t mAudioSamplingFrequency = 0;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.6 - ControlDataCommon::ConnectionStatus

ControlDataCommon::ConnectionStatus

Connection status struct containing information about a connection event.

#include <ControlDataCommon.h>

Public Functions

Public Attributes

Public Functions Documentation

function ConnectionStatus

ConnectionStatus() =default

function ConnectionStatus

inline ConnectionStatus(
    ConnectionType mConnectionType,
    const std::string & mIp,
    uint16_t mPort
)

Public Attributes Documentation

variable mConnectionType

ConnectionType mConnectionType = ConnectionType::DISCONNECTED;

variable mIP

std::string mIP;

variable mPort

uint16_t mPort = 0;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.7 - ControlDataCommon::Response

ControlDataCommon::Response

A response from a ControlDataReceiver to a request. The UUID tells which receiver the response is sent from.

#include <ControlDataCommon.h>

Public Attributes

Public Attributes Documentation

variable mMessage

std::vector< uint8_t > mMessage;

variable mRequestId

uint64_t mRequestId = 0;

The actual message.

variable mFromUUID

std::string mFromUUID;

The ID of the request this is a response to.

Updated on 2024-04-08 at 15:15:27 +0200

4.1.8 - ControlDataCommon::StatusMessage

ControlDataCommon::StatusMessage

A status message from a ControlDataReceiver. The UUID tells which receiver the message is sent from.

#include <ControlDataCommon.h>

Public Attributes

Public Attributes Documentation

variable mMessage

std::vector< uint8_t > mMessage;

variable mFromUUID

std::string mFromUUID;

The actual message.

Updated on 2024-04-08 at 15:15:27 +0200

4.1.9 - ControlDataSender

ControlDataSender

A ControlDataSender can send control signals to one or more receivers using a network connection. A single ControlDataSender can connect to multiple receivers, all identified by a UUID. The class is controlled using an ISystemControllerInterface; this interface is responsible for setting up connections to receivers. The ControlDataSender can send asynchronous requests to (all) the receivers and get a response back. Each response is identified with a request ID as well as the UUID of the responding receiver. The ControlDataSender can also receive status messages from the receivers.

#include <ControlDataSender.h>

Public Classes

Public Functions

Public Functions Documentation

function ControlDataSender

ControlDataSender()

Default constructor, creates an empty object.

function ~ControlDataSender

~ControlDataSender()

Destructor. Will disconnect from the connected receivers and close the System controller connection.

function configure

bool configure(
    const std::shared_ptr< ISystemControllerInterface > & controllerInterface,
    const Settings & settings
)

Configure this ControlDataSender and connect it to the System Controller. This method will fail in case the ISystemControllerInterface has already been connected to the controller by another component, as such interface can only be used by one component.

Parameters:

• controllerInterface The interface to the System controller, used for communicating with this ControlDataSender
• settings The Settings to use for this ControlDataSender

Return: True on success, false otherwise

function sendRequestToReceivers

bool sendRequestToReceivers(
    const std::vector< uint8_t > & request,
    uint64_t & requestId
)

Send a request to all the connected ControlDataReceivers asynchronously. The responses will be sent to the response callback.

Parameters:

• request The request message
• requestId The unique identifier of this request. Used to identify the async responses.

Return: True if the request was successfully sent, false otherwise

function sendMultiRequestToReceivers

bool sendMultiRequestToReceivers(
    const std::vector< std::vector< uint8_t >> & request,
    uint64_t & requestId
)

Send a multi-message request to all the connected ControlDataReceivers asynchronously. The responses will be sent to the response callback.

Parameters:

• request The request message
• requestId The unique identifier of this request. Used to identify the async responses.

Return: True if the request was successfully sent, false otherwise

function ControlDataSender

ControlDataSender(
    ControlDataSender const & 
) =delete

function operator=

ControlDataSender & operator=(
    ControlDataSender const & 
) =delete

function getVersion

static std::string getVersion()

Get application version.

Return: a string with the current version, e.g. “6.0.0-39-g60a35937”

Updated on 2024-04-08 at 15:15:27 +0200

4.1.10 - ControlDataSender::Settings

ControlDataSender::Settings

Settings for a ControlDataSender.

#include <ControlDataSender.h>

Public Attributes

Public Attributes Documentation

variable mConnectionStatusCallback

std::function< void(const ControlDataCommon::ConnectionStatus &)> mConnectionStatusCallback;

variable mResponseCallback

std::function< void(const ControlDataCommon::Response &)> mResponseCallback;

variable mStatusMessageCallback

std::function< void(const ControlDataCommon::StatusMessage &)> mStatusMessageCallback;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.11 - DeviceMemory

DeviceMemory

RAII class for a CUDA memory buffer.

#include <DeviceMemory.h>

Public Functions

Public Functions Documentation

function DeviceMemory

DeviceMemory() =default

Default constructor, creates an empty object, without allocating any memory on the device.

function DeviceMemory

explicit DeviceMemory(
    size_t numberOfBytes
)

Constructor allocating the required number of bytes.

Parameters:

• numberOfBytes Number of bytes to allocate

Exceptions:

• std::runtime_error In case the allocation failed

function DeviceMemory

explicit DeviceMemory(
    size_t numberOfBytes,
    cudaStream_t cudaStream
)

Constructor allocating the required number of bytes by making an async allocation. The allocation will be put in the given CUDA stream.

Parameters:

• size Number of bytes to allocate
• cudaStream The CUDA stream to put the async allocation in. A reference to this stream will also be saved internally to be used to asynchronously free them memory when the instance goes out of scope.

Exceptions:

• std::runtime_error In case the async allocation failed to be put in queue.

See: freeMemoryAsync method is not explicitly called, the memory will be freed synchronously when this DeviceMemory instance goes out of scope, meaning that the entire GPU is synchronized, which will impact performance negatively.

Note:

• The method will return as soon as the allocation is put in queue in the CUDA stream, i.e. before the actual allocation is made. Using this DeviceMemory is only valid as long as it is used in the same CUDA stream, or in case another stream is used, only if that stream is synchronized first with respect to cudaStream.
• In case the

function DeviceMemory

explicit DeviceMemory(
    void * deviceMemory
)

Constructor taking ownership of an already allocated CUDA memory pointer. This class will free the pointer once it goes out of scope.

Parameters:

• deviceMemory CUDA memory pointer to take ownership over.

function allocateMemory

bool allocateMemory(
    size_t numberOfBytes
)

Allocates device memory. The memory allocated will automatically be freed by the destructor.

Parameters:

• numberOfBytes Number of bytes to allocate

Return: True on success, false if there is already memory allocated by this instance, or if the CUDA malloc failed.

function allocateMemoryAsync

bool allocateMemoryAsync(
    size_t numberOfBytes,
    cudaStream_t cudaStream
)

Allocates device memory. The memory allocated will automatically be freed by the destructor.

Parameters:

• numberOfBytes Number of bytes to allocate
• cudaStream The CUDA stream to use for the allocation. A reference to this stream will also be saved internally to be used to asynchronously free them memory when the instance goes out of scope.

Return: True on success, false if there is already memory allocated by this instance, or if the CUDA malloc failed.

function reallocateMemory

bool reallocateMemory(
    size_t numberOfBytes
)

Reallocates device memory. Already existing memory allocation will be freed before the new allocation is made. In case this DeviceMemory has no earlier memory allocation, this method will just allocate new CUDA memory and return a success status.

Parameters:

• numberOfBytes Number of bytes to allocate in the new allocation

Return: True on success, false if CUDA free or CUDA malloc failed.

function reallocateMemoryAsync

bool reallocateMemoryAsync(
    size_t numberOfBytes,
    cudaStream_t cudaStream
)

Asynchronously reallocate device memory. Already existing memory allocation will be freed before the new allocation is made. In case this DeviceMemory has no earlier memory allocation, this method will just allocate new CUDA memory and return a success status.

Parameters:

• numberOfBytes Number of bytes to allocate in the new allocation
• cudaStream The CUDA stream to use for the allocation and freeing of memory. A reference to this stream will also be saved internally to be used to asynchronously free them memory when this instance goes out of scope.

Return: True on success, false if CUDA free or CUDA malloc failed.

function allocateAndResetMemory

bool allocateAndResetMemory(
    size_t numberOfBytes
)

Allocates device memory and resets all bytes to zeroes. The memory allocated will automatically be freed by the destructor.

Parameters:

• numberOfBytes Number of bytes to allocate

Return: True on success, false if there is already memory allocated by this instance, or if any of the CUDA operations failed.

function allocateAndResetMemoryAsync

bool allocateAndResetMemoryAsync(
    size_t numberOfBytes,
    cudaStream_t cudaStream
)

Allocates device memory and resets all bytes to zeroes. The memory allocated will automatically be freed by the destructor.

Parameters:

• numberOfBytes Number of bytes to allocate
• cudaStream The CUDA stream to use for the allocation and resetting. A reference to this stream will also be saved internally to be used to asynchronously free them memory when the instance goes out of scope.

Return: True on success, false if there is already memory allocated by this instance, or if any of the CUDA operations failed.

function freeMemory

bool freeMemory()

Free the device memory held by this class. Calling this when no memory is allocated is a no-op.

See: freeMemoryAsync instead.

Return: True in case the memory was successfully freed (or not allocated to begin with), false otherwise.

Note: This method will free the memory in an synchronous fashion, synchronizing the entire CUDA context and ignoring the internally saved CUDA stream reference in case one exist. For async freeing of the memory, use

function freeMemoryAsync

bool freeMemoryAsync(
    cudaStream_t cudaStream
)

Deallocate memory asynchronously, in a given CUDA stream. Calling this when no memory is allocated is a no-op.

Parameters:

• cudaStream The CUDA stream to free the memory asynchronously in

Return: True in case the memory deallocation request was successfully put in queue in the CUDA stream.

Note:

• The method will return as soon as the deallocation is put in queue in the CUDA stream, i.e. before the actual deallocation is made.
• It is the programmer’s responsibility to ensure this DeviceMemory is not used in another CUDA stream before this method is called. In case it is used in another CUDA stream, sufficient synchronization must be made before calling this method (and a very good reason given for not freeing the memory in that CUDA stream instead)

function setFreeingCudaStream

void setFreeingCudaStream(
    cudaStream_t cudaStream
)

Set which CUDA stream to use for freeing this DeviceMemory. In case the DeviceMemory already holds a CUDA stream to use for freeing the memory, this will be overwritten.

Parameters:

• cudaStream The new CUDA stream to use for freeing the memory when the destructor is called.

Note: It is the programmer’s responsibility to ensure this DeviceMemory is not used in another CUDA stream before this instance is destructed. In case it is used in another CUDA stream, sufficient synchronization must be made before setting this as the new CUDA stream to use when freeing the memory.

function ~DeviceMemory

~DeviceMemory()

Destructor, frees the internal CUDA memory.

function getDevicePointer

template <typename T  =uint8_t>
inline T * getDevicePointer() const

Template Parameters:

• T The pointer type to return

Return: the CUDA memory pointer handled by this class. Nullptr in case no memory is allocated.

function getSize

size_t getSize() const

Return: The size of the CUDA memory allocation held by this class.

function DeviceMemory

DeviceMemory(
    DeviceMemory && other
)

function operator=

DeviceMemory & operator=(
    DeviceMemory && other
)

function swap

void swap(
    DeviceMemory & other
)

function DeviceMemory

DeviceMemory(
    DeviceMemory const & 
) =delete

DeviceMemory is not copyable.

function operator=

DeviceMemory operator=(
    DeviceMemory const & 
) =delete

Updated on 2024-04-08 at 15:15:27 +0200

4.1.12 - IControlDataReceiver

IControlDataReceiver

IControlDataReceiver is the interface class for the control data receiver. An IControlDataReceiver can receive messages from a sender or other IControlDataReceivers using a network connection. It can also connect to and forward the incoming request messages to other receivers. The connections to the sender and the other receivers are controlled by an ISystemControllerInterface instance. The ControlDataReceiver has a receiving or listening side, as well as a sending side. The listening side can listen to one single network port and have multiple ControlDataSenders and ControlDataReceivers connected to that port to receive requests from them. On the sending side of the ControlDataReceiver, it can be connected to the listening side of other ControlDataReceivers, used to forward all incoming messages to that receiver, as well as sending its own requests. More…

#include <IControlDataReceiver.h>

Public Classes

Public Functions

Detailed Description

class IControlDataReceiver;

IControlDataReceiver is the interface class for the control data receiver. An IControlDataReceiver can receive messages from a sender or other IControlDataReceivers using a network connection. It can also connect to and forward the incoming request messages to other receivers. The connections to the sender and the other receivers are controlled by an ISystemControllerInterface instance. The ControlDataReceiver has a receiving or listening side, as well as a sending side. The listening side can listen to one single network port and have multiple ControlDataSenders and ControlDataReceivers connected to that port to receive requests from them. On the sending side of the ControlDataReceiver, it can be connected to the listening side of other ControlDataReceivers, used to forward all incoming messages to that receiver, as well as sending its own requests.

Each ControlDataReceiver can be configured to have a certain message delay. This delay parameter is set when the System controller instructs the ControlDataReceiver to start listening for incoming connections from senders (or sending ControlDataReceiver). Each incoming request will then be delayed and delivered according to the parameter, compared to the send timestamp in the message. In case multiple receivers are chained, like sender->receiver1->receiver2, receiver1 will delay the incoming messages from the sender based on when they were sent from the sender. Furthermore, receiver2 will delay them compared to when they were sent from receiver1.

An IControlDataReceiver can send status messages back to the sender. In case of chained receivers, the message will be forwarded back to the sender. A user of the ControlDataReceiver can register callbacks to receive requests and status messages. There is also an optional “preview” callback that is useful in case the incoming messages are delayed (have a delay > 0). This callback will then be called as soon as the request message arrives, to allow the user to prepare for when the actual delayed request callback is called.

Public Functions Documentation

function ~IControlDataReceiver

virtual ~IControlDataReceiver() =default

Destructor.

function configure

virtual bool configure(
    const Settings & settings
) =0

Configure this instance.

Parameters:

• settings The settings to use for this receiver

Return: True on success, false otherwise

function sendStatusMessageToSender

virtual bool sendStatusMessageToSender(
    const std::vector< uint8_t > & message
) =0

Send a status message to the (directly or indirectly) connected ControlDataSender. In case this ControlDataReceiver has another ControlDataReceiver as sender, that receiver will forward the status message to the sender.

Parameters:

• message The status message

Return: True in case the message was sent successfully, false otherwise.

function sendRequestToReceivers

virtual bool sendRequestToReceivers(
    const std::vector< uint8_t > & request,
    uint64_t & requestId
) =0

Send a request to the connected ControlDataReceivers asynchronously. This request will only be sent to ControlDataReceivers on the sending side of this receiver. In case a receiver is located between this ControlDataReceiver and the sender, neither of those will see this request. The response will be sent to the response callback asynchronously.

Parameters:

• request The request message
• requestId The unique identifier of this request. Used to identify the async response.

Return: True if the request was successfully sent, false otherwise

function sendMultiRequestToReceivers

virtual bool sendMultiRequestToReceivers(
    const std::vector< std::vector< uint8_t >> & request,
    uint64_t & requestId
) =0

Send a multi-message request to the connected ControlDataReceivers asynchronously. This request will only be sent to ControlDataReceivers on the sending side of this receiver. In case a receiver is located between this ControlDataReceiver and the sender, neither of those will see this request. The response will be sent to the response callback asynchronously.

Parameters:

• request The request message
• requestId The unique identifier of this request. Used to identify the async response.

Return: True if the request was successfully sent, false otherwise

Updated on 2024-04-08 at 15:15:27 +0200

4.1.13 - IControlDataReceiver::IncomingRequest

IControlDataReceiver::IncomingRequest

An incoming request to this ControlDataReceiver.

#include <IControlDataReceiver.h>

Public Attributes

Public Attributes Documentation

variable mMessages

std::vector< std::vector< uint8_t > > mMessages;

variable mSenderUUID

std::string mSenderUUID;

The actual messages.

variable mRequesterUUID

std::string mRequesterUUID;

UUID of the sender/forwarder that sent the request to this ControlDataReceiver.

variable mRequestID

uint64_t mRequestID = 0;

UUID of the requester (the original sender, creating the request)

variable mSenderTimestampUs

int64_t mSenderTimestampUs =
            0;

The requester’s unique id of this request.

variable mDeliveryTimestampUs

int64_t mDeliveryTimestampUs =
            0;

The TAI timestamp when this message was sent from the sender/forwarder in micro sec since TAI epoch.

Updated on 2024-04-08 at 15:15:27 +0200

4.1.14 - IControlDataReceiver::ReceiverResponse

IControlDataReceiver::ReceiverResponse

A response message to a request.

#include <IControlDataReceiver.h>

Public Attributes

Public Attributes Documentation

variable mMessage

std::vector< uint8_t > mMessage;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.15 - IControlDataReceiver::Settings

IControlDataReceiver::Settings

Settings for a ControlDataReceiver.

#include <IControlDataReceiver.h>

Public Attributes

Public Attributes Documentation

variable mProductionPipelineUUID

std::string mProductionPipelineUUID;

variable mPreviewIncomingRequestCallback

std::function< void(const IncomingRequest &)> mPreviewIncomingRequestCallback;

UUID of the Production Pipeline component this Receiver is a part of Callback called as soon as this ControlDataReceiver gets the request, as a “preview” of what requests will be delivered.

variable mIncomingRequestCallback

std::function< ReceiverResponse(const IncomingRequest &)> mIncomingRequestCallback;

The actual callback used when this ControlDataReceiver gets a request and wants a response back. This callback will delay and deliver the request according to this receiver’s configured delay.

variable mConnectionStatusCallback

std::function< void(const ControlDataCommon::ConnectionStatus &)> mConnectionStatusCallback;

variable mResponseCallback

std::function< void(const ControlDataCommon::Response &)> mResponseCallback;

Callback for connection events.

Updated on 2024-04-08 at 15:15:27 +0200

4.1.16 - IMediaStreamer

IMediaStreamer

IMediaStreamer is an interface class for MediaStreamers, that can take a single stream of uncompressed video and/or audio frames and encode and output it in some way. This output can either be a stream to a network or writing down the data to a file on the hard drive. This class is configured from two interfaces. The input configuration (input video resolution, frame rate, pixel format, number of audio channels…) is made through this C++ API. The output stream is then started from the System Controller. Any of these configurations can be made first. The actual stream to output will start once the first call to. More…

#include <IMediaStreamer.h>

Public Classes

Public Functions

Detailed Description

class IMediaStreamer;

IMediaStreamer is an interface class for MediaStreamers, that can take a single stream of uncompressed video and/or audio frames and encode and output it in some way. This output can either be a stream to a network or writing down the data to a file on the hard drive. This class is configured from two interfaces. The input configuration (input video resolution, frame rate, pixel format, number of audio channels…) is made through this C++ API. The output stream is then started from the System Controller. Any of these configurations can be made first. The actual stream to output will start once the first call to.

See: outputData is made.

Public Functions Documentation

function ~IMediaStreamer

virtual ~IMediaStreamer() =default

Destructor.

function configure

virtual bool configure(
    const std::string & uuid,
    const Settings & settings,
    CUcontext cudaContext
) =0

Configure this MediaStreamer. This must be called before any call to the.

Parameters:

• uuid UUID of this MediaStreamer
• settings Settings for this MediaStreamer
• cudaContext The CUDA context to use for this MediaStreamer. The frames passed to this instance should be valid in this CUDA context. The streamer will use this context for preprocessing and encoding.

See:

• setInputFormatAndStart or
• outputData

Return: True if the streamer was successfully configured

function setInputFormatAndStart

virtual bool setInputFormatAndStart(
    const Configuration & configuration
) =0

Set the input format of this MediaStreamer and start the streamer. The.

Parameters:

• configuration The configuration with the format of the frames that will be sent to this MediaReceiver

See:

• configure method must be called before this method is called. This method must be called before any call to
• outputData. If the format should be reset, the
• stopAndResetFormat method should be called first and then this method can be called again to reset the format.

Return: True if the streamer was successfully started, false otherwise

function stopAndResetFormat

virtual bool stopAndResetFormat() =0

Stop streaming and reset the format. A call to this method will stop any output streams set up by the ISystemControllerInterface and reset the input format set by the.

See: setInputFormatAndStart method. The connection to the ISystemControllerInterface will be kept.

Return: True if the stream was successfully stopped and the format reset, or if the format was not set before this method was called, false on error.

function hasFormatAndIsRunning

virtual bool hasFormatAndIsRunning() const =0

Return: True if the input format is set and the OutputStreamer is running, false otherwise

function hasOpenOutputStream

virtual bool hasOpenOutputStream() const =0

See:

• outputData will be discarded without being encoded, as there is no stream to output them to. This method can be used to check if the frame even needs to be produced by the rendering engine. Note however, that the
• outputData method will still log the frames sent to it as received, even if they are not encoded when the output stream is closed.

Return: True if the output stream of the MediaStreamer is currently open and outputting the data, false otherwise. In case this returns false all frames passed to

function outputData

virtual bool outputData(
    const AlignedFramePtr & frame
) =0

Output data through this streamer. The AlignedFrame::mRenderingTimestamp of the frame will be used as PTS when encoding the uncompressed frame.

Parameters:

• frame The data frame to output, with video data in CUDA memory

See: setInputFormatAndStart method, or if the format has not been set by a call to that method.

Return: True if the frame was accepted (but not necessarily streamed, in case the output stream has not been set up by the ISystemControllerInterface), false in case the frame did not match the configuration made in the

Note: The DeviceMemory in the AlignedFrame passed to this method must not be in use by any CUDA stream. In case the memory has been used in kernels in another CUDA stream, make sure to first synchronize with the stream, before passing it over to the MediaStreamer.

Updated on 2024-04-08 at 15:15:27 +0200

4.1.17 - IMediaStreamer::Configuration

IMediaStreamer::Configuration

The input configuration of the frames that will be sent to this MediaStreamer. The output stream configuration is made from the System controller via the ISystemControllerInterface.

#include <IMediaStreamer.h>

Public Attributes

Public Attributes Documentation

variable mIncomingPixelFormat

PixelFormat mIncomingPixelFormat = PixelFormat::kUnknown;

variable mWidth

uint32_t mWidth = 0;

variable mHeight

uint32_t mHeight = 0;

variable mFrameRateN

uint32_t mFrameRateN = 0;

variable mFrameRateD

uint32_t mFrameRateD = 0;

variable mAudioSampleRate

uint32_t mAudioSampleRate = 0;

variable mNumAudioChannels

uint32_t mNumAudioChannels = 0;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.18 - IMediaStreamer::Settings

IMediaStreamer::Settings

Settings used when creating a new MediaStreamer.

#include <IMediaStreamer.h>

Public Attributes

Public Attributes Documentation

variable mName

std::string mName;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.19 - IngestApplication

IngestApplication

Public Classes

Public Functions

Public Functions Documentation

function IngestApplication

IngestApplication()

Constructor, creates an empty IngestApplication without starting it.

function ~IngestApplication

~IngestApplication()

Destructor.

function start

bool start(
    const std::shared_ptr< ISystemControllerInterface > & controllerInterface,
    const Settings & settings
)

Start this IngestApplication given an interface to the System Controller and an UUID.

Parameters:

• controllerInterface The interface for communication with the System Controller
• settings The settings for the IngestApplication

Return: True if the IngestApplication was successfully started, false otherwise

function stop

bool stop()

Stop this IngestApplication.

Return: True if the IngestApplication was successfully stopped, false otherwise

function getVersion

static std::string getVersion()

Get application version.

Return: a string with the current version, e.g. “6.0.0-39-g60a35937”

function getLibraryVersions

static std::string getLibraryVersions()

Get the versions of the libraries available at runtime, among others, CUDA version, BMD and NDI versions.

Return: a string with the currently found versions of the libraries used by this application

Updated on 2024-04-08 at 15:15:27 +0200

4.1.20 - IngestApplication::Settings

IngestApplication::Settings

Updated on 2024-04-08 at 15:15:27 +0200

4.1.21 - ISystemControllerInterface

ISystemControllerInterface

An ISystemControllerInterface is the interface between a component and the System controller controlling the component. The interface allows for two-way communication between the component and the system controller by means of sending requests and getting responses. Classes deriving from the ISystemControllerInterface should provide the component side implementation of the communication with the system controller. This interface can be inherited and implemented by developers to connect to custom system controllers, or to directly control a component programmatically, without the need for connecting to a remote server.

#include <ISystemControllerInterface.h>

Inherited by SystemControllerConnection

Public Classes

Public Types

Public Functions

Public Types Documentation

enum StatusCode

Status codes used in JSON response messages for Websockets. These are starting at 3000 since the 1000 - 2000 range is taken up by the Spec: https://datatracker.ietf.org/doc/html/rfc6455#section-7.4.1.

Public Functions Documentation

function ~ISystemControllerInterface

virtual ~ISystemControllerInterface() =default

Virtual destructor.

function sendMessage

virtual std::optional< std::string > sendMessage(
    const std::string & messageTitle,
    const nlohmann::json & parameters
) =0

Send a message containing a JSON object to the controller.

Parameters:

• messageTitle The title of the status type or request
• parameters The parameters part of the JSON message

Return: Optional containing an error message on error, else nullopt in case the message was successfully sent

Reimplemented by: SystemControllerConnection::sendMessage

function registerRequestCallback

virtual bool registerRequestCallback(
    const Callbacks & callbacks
) =0

Register the callbacks to call for events in this class.

Parameters:

• callbacks The callbacks to use when events in this class happen

Return: True on successful registration, false if some callback is not set or if already connected

Reimplemented by: SystemControllerConnection::registerRequestCallback

function connect

virtual bool connect() =0

Connect to the System controller.

Return: True on successful connection, false on error or if already connected

Reimplemented by: SystemControllerConnection::connect

function disconnect

virtual bool disconnect() =0

Disconnect from the System controller.

Return: True on successful disconnection, false on error or if not connected

Reimplemented by: SystemControllerConnection::disconnect

function isConnected

virtual bool isConnected() const =0

Return: True if connected to the System controller, false otherwise

Reimplemented by: SystemControllerConnection::isConnected

function getUUID

virtual std::string getUUID() const =0

Return: The UUID of this interface to the System controller

Reimplemented by: SystemControllerConnection::getUUID

Updated on 2024-04-08 at 15:15:27 +0200

4.1.22 - ISystemControllerInterface::Callbacks

ISystemControllerInterface::Callbacks

A struct containing the callbacks that needs to be registered by the component using this interface.

#include <ISystemControllerInterface.h>

Public Attributes

Public Attributes Documentation

variable mRequestCallback

std::function< Response(const std::string &, const nlohmann::json &)> mRequestCallback;

variable mConnectionClosedCallback

std::function< void(uint32_t, const std::string &, const std::error_code &)> mConnectionClosedCallback;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.23 - ISystemControllerInterface::Response

ISystemControllerInterface::Response

A response to a request, consists of a status code and an (optional) parameters JSON object.

#include <ISystemControllerInterface.h>

Public Attributes

Public Attributes Documentation

variable mCode

StatusCode mCode;

variable mParameters

nlohmann::json mParameters;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.24 - MediaReceiver

MediaReceiver

A MediaReceiver contains the logic for receiving, decoding and aligning incoming media sources from the Ingests. The aligned data is then delivered to the Rendering Engine which is also responsible for setting up the MediaReceiver. The MediaReceiver has a builtin multi view generator, which can create output streams containing composited subsets of the incoming video sources. This class is controlled using an ISystemControllerInterface provided when starting it.

#include <MediaReceiver.h>

Public Classes

Public Types

Public Functions

Public Types Documentation

enum TallyBorderColor

Available colors for tally border in multi view.

Public Functions Documentation

function MediaReceiver

MediaReceiver()

Default constructor.

function ~MediaReceiver

~MediaReceiver()

Default destructor.

function start

bool start(
    const std::shared_ptr< ISystemControllerInterface > & controllerInterface,
    CUcontext cudaContext,
    const Settings & settings,
    const ControlDataReceiver::Settings & receiverSettings
)

Start the MediaReceiver. This method will call connect on the System controller interface and set up the callbacks from the interface to call internal methods.

Parameters:

• controllerInterface The ISystemControllerInterface to use for this MediaReceiver. The interface should be configured (Such as setting the IP address and port of the System Controller if a server based System Controller is used) but not connected before passed to this method. This method will internally set the callbacks before connecting to the controller. If the controller is already connected or if the controller is not configured, this method will return false. This class will take ownership of the smart pointer.
• cudaContext The CUDA context to use for this MediaReceiver. The frames will delivered as CUDA pointers, valid in this context, and the multi-view generator will use this context for rendering and encoding.
• settings The settings to use for the MediaReceiver.
• receiverSettings The settings to use for the ControlDataReceiver.

Return: True if the MediaReceiver was started successfully, false otherwise.

function stop

void stop()

Stop the MediaReceiver.

function getCustomMultiViewSourceInput

std::function< void(const AlignedFramePtr &)> getCustomMultiViewSourceInput(
    uint32_t inputSlot,
    const std::string & name =""
)

This method allows the Rendering Engine to provide custom input sources to the Multi-view generator to send video streams that can be added to the multi-views. This could for instance be used for adding a “preview” of the video stream the rendering engine is about to cut to.

Parameters:

• inputSlot The input slot this source will be “plugged in” to. The custom input sources share the input slots with the streams connected from Ingests. This means that it is probably a good idea to use higher numbered slots for these custom inputs, such as numbers from 1000, so that the lower numbers, 1 and up, can be used by the connected video cameras, as the input slot number will also be used when cutting.
• name Optional human readable name of this stream, to be presented to the System Controller.

See: MediaReceiver::Settings::mDecodedFormat.

Return: A function to where the Rendering Engine should send the frames. In case the requested inputSlot is already used by another custom input source, or a stream from an ingest, the returned function will be nullptr.

Note: Make sure no CUDA stream will write to the DeviceMemory in the AlignedFrame passed to the function. Failing to do so will lead to undefined behavior. In case another CUDA stream has written to the DeviceMemory, make sure to synchronize with the stream before passing the AlignedFrame.

Precondition: The AlignedFrame sent to this function must have the same pixel format as this MediaReceiver is configured to deliver to the Rendering Engine,

function removeCustomMultiViewSourceInput

bool removeCustomMultiViewSourceInput(
    uint32_t inputSlot
)

Remove a custom multi-view generator source input earlier registered using the getCustomMultiViewSourceInput method.

Parameters:

• inputSlot The input slot the custom multi-view source input is connected to, that should be removed

Return: True if the custom multi-view source input was successfully removed, false in case there was no custom input registered for the given input slot, or in case of an internal error.

function createMediaStreamerOutput

std::shared_ptr< IMediaStreamer > createMediaStreamerOutput(
    const MediaStreamer::Settings & settings
)

Create a new MediaStreamer instance to output data from this MediaReceiver.

Parameters:

• settings The settings for the new MediaStreamer

Return: A shared pointer to the MediaStreamer instance in case of a successful creation, otherwise nullptr

function removeMediaStreamerOutput

bool removeMediaStreamerOutput(
    const std::string & uuid
)

Remove an MediaStreamer created via the.

Parameters:

• uuid The UUID of the MediaStreamer to remove

See: createMediaStreamerOutput method by its UUID.

Return: True if the MediaStreamer was successfully removed, false in case no MediaStreamer with the given UUID was found.

function getControlDataReceiver

std::shared_ptr< IControlDataReceiver > getControlDataReceiver()

Get a pointer to the ControlDataReceiver instance of this MediaReceiver. A call to this method will always return the same instance.

See: start.

Return: Pointer to the ControlDataReceiver instance. Nullptr if called before a successful call to

function setTallyBorder

void setTallyBorder(
    uint32_t inputSlot,
    TallyBorderColor color
)

Set tally border color in the multi-views for a specific input slot.

Parameters:

• inputSlot the input slot for a source
• color the color to set

function clearTallyBorder

void clearTallyBorder(
    uint32_t inputSlot
)

Remove tally border in the multi-views for a specific input slot.

Parameters:

• inputSlot the input slot for a source

function clearAllTallyBorders

void clearAllTallyBorders()

Remove all tally borders.

function getTallyBorder

MediaReceiver::TallyBorderColor getTallyBorder(
    uint32_t inputSlot
) const

get tally border color for an input slot

Parameters:

• inputSlot the input slot to get color for

Return: the tally border color for the given input slot

function MediaReceiver

MediaReceiver(
    MediaReceiver const & 
) =delete

MediaReceiver is neither copyable nor movable.

function MediaReceiver

MediaReceiver(
    MediaReceiver && 
) =delete

function operator=

MediaReceiver & operator=(
    MediaReceiver const & 
) =delete

function operator=

MediaReceiver & operator=(
    MediaReceiver && 
) =delete

function getVersion

static std::string getVersion()

Get application version.

Return: a string with the current version, e.g. “6.0.0-39-g60a35937”

function getLibraryVersions

static std::string getLibraryVersions()

Get versions of the libraries available at runtime, among others, CUDA runtime and driver versions.

Return: a string with the currently found versions of the libraries used by this application

Updated on 2024-04-08 at 15:15:27 +0200

4.1.25 - MediaReceiver::NewStreamParameters

MediaReceiver::NewStreamParameters

A struct containing information on the format of an incoming stream.

#include <MediaReceiver.h>

Public Attributes

Public Attributes Documentation

variable mVideoHeight

uint32_t mVideoHeight = 0;

variable mVideoWidth

uint32_t mVideoWidth = 0;

variable mFrameRateN

uint32_t mFrameRateN = 0;

variable mFrameRateD

uint32_t mFrameRateD = 1;

variable mAudioSampleRate

uint32_t mAudioSampleRate = 0;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.26 - MediaReceiver::Settings

MediaReceiver::Settings

Settings for a MediaReceiver.

#include <MediaReceiver.h>

Public Attributes

Public Attributes Documentation

variable mDecodedFormat

PixelFormat mDecodedFormat = PixelFormat::kRgba64Le;

The pixel format delivered to the rendering engine.

variable mNewConnectionCallback

std::function< std::function< void(const AlignedFramePtr &)>uint32_t inputSlot, const std::string &streamID, const NewStreamParameters &newStreamParameters)> mNewConnectionCallback;

Parameters:

• inputSlot can be seen as a virtual SDI input on a hardware mixer and should be used by the Rendering engine to identify sources. For example, if a source is connected to input slot 5, the button “Cut to camera 5” on the control panel ought to cut to this stream. The MediaReceiver is responsible for making sure only one stream can be connected to an input slot at a time. This callback might be called multiple times with the same input slot, however, in case the stream has been disconnected and the mClosedConnectionCallback has been called for that input slot earlier.
• streamID is the identification string of the video/audio source on the Ingest. A Rendering engine might have the same StreamID connected multiple times (for instance if two different alignments are used for the same source) so this should not be used as an unique identifier for the stream.
• newStreamParameters contains information on the pending incoming stream to allow the Rendering engine to decide if it can receive it or not.

Note: The internal CUDA stream is synchronized before delivering the AlignedFrames, meaning the DeviceMemory in the AlignedFrame will always contain the written data; no further synchronization is needed before the rendering engine can read the data.

A callback called by the MediaReceiver whenever a new connection from an Ingest is set up. The callback should return a function to which the data will be delivered. In case the Rendering engine does not want to accept the pending incoming stream (due to a format not supported, etc) the Rendering engine can return nullptr.

variable mClosedConnectionCallback

std::function< void(uint32_t inputSlot)> mClosedConnectionCallback;

Parameters:

• inputSlot The input slot that was disconnected

A callback called whenever a connection from an Ingest has been stopped. When receiving this callback, the callback function returned by the mNewConnectionCallback will not be called anymore and can be discarded if required by the application. After this function has been called, the input slot might be reused by another incoming stream after a call to the mNewConnectionCallback.

variable mResetCallback

std::function< std::optional< std::string >)> mResetCallback;

See: mNewConnectionCallback should be kept, however).

Return: True on successful reset, false in case something unexpected happened

A callback called whenever the System Controller has requested that the Rendering Engine should be reset to its initial state (connections made via the

variable mUseMultiViewer

bool mUseMultiViewer = false;

variable mDeliverOld

bool mDeliverOld = false;

Set to true if this MediaReceiver should have a multi-view generator.

variable mAlignedFrameFreeStream

CUstream mAlignedFrameFreeStream = nullptr;

Set to true if the media frames should be delivered, even if the delivery time has already passed, otherwise they are discarded The CUDA stream to use for asynchronously freeing the AlignedFrames’ DeviceMemory. Set this to the rendering engine’s CUDA stream, to ensure memory is not freed before the rendering engine is done using it asynchronously in its CUDA stream. Note that the MediaReceiver instance and all AlignedFrames delivered from it must be destroyed before the CUDA stream can be destroyed.

Updated on 2024-04-08 at 15:15:27 +0200

4.1.27 - SystemControllerConnection

SystemControllerConnection

An implementation of the ISystemControllerInterface for a System controller residing in a remote server. The connection to the server uses a Websocket.

#include <SystemControllerConnection.h>

Inherits from ISystemControllerInterface

Public Classes

Public Types

Public Functions

Additional inherited members

Public Classes inherited from ISystemControllerInterface

Public Types inherited from ISystemControllerInterface

Public Functions inherited from ISystemControllerInterface

Public Types Documentation

enum ComponentType

Enumeration of component types the component using this SystemControllerConnection can tell the System Controller to be seen as.

Public Functions Documentation

function SystemControllerConnection

SystemControllerConnection()

function ~SystemControllerConnection

~SystemControllerConnection() override

function configure

bool configure(
    const Settings & settings
)

Configure this connection. This method should be called before calling.

Parameters:

• settings The settings to use when connecting to the server

See: connect.

Return: True if successfully configured, false otherwise

function connect

virtual bool connect() override

Connect to the server using the settings set with the.

See: configure method.

Return: True if connection was successful, false otherwise

Reimplements: ISystemControllerInterface::connect

function isConnected

virtual bool isConnected() const override

Return: True if this class is connected to the server, false otherwise

Reimplements: ISystemControllerInterface::isConnected

function getUUID

virtual std::string getUUID() const override

Return: The UUID of this interface to the System controller

Reimplements: ISystemControllerInterface::getUUID

function sendMessage

virtual std::optional< std::string > sendMessage(
    const std::string & messageTitle,
    const nlohmann::json & parameters
) override

Send a message containing a JSON object to the system controller server.

Parameters:

• messageTitle The title of the status type or request
• parameters The parameters part of the JSON message

Return: Optional containing an error message on error, else nullopt in case the message was successfully sent

Reimplements: ISystemControllerInterface::sendMessage

function disconnect

virtual bool disconnect() override

Disconnect from the server.

Return: True if successfully disconnected, false on internal error

Reimplements: ISystemControllerInterface::disconnect

function registerRequestCallback

virtual bool registerRequestCallback(
    const Callbacks & callbacks
) override

Register callbacks to call when getting requests from the server or the server connection is lost.

Parameters:

• callbacks The callbacks to set

Return: True if successfully registered, false if a callback is not set, or if already connected to the server

Reimplements: ISystemControllerInterface::registerRequestCallback

function SystemControllerConnection

SystemControllerConnection(
    SystemControllerConnection const & 
) =delete

function SystemControllerConnection

SystemControllerConnection(
    SystemControllerConnection && 
) =delete

function operator=

SystemControllerConnection & operator=(
    SystemControllerConnection const & 
) =delete

function operator=

SystemControllerConnection & operator=(
    SystemControllerConnection && 
) =delete

Updated on 2024-04-08 at 15:15:27 +0200

4.1.28 - SystemControllerConnection::Settings

SystemControllerConnection::Settings

Settings for a SystemControllerConnection.

#include <SystemControllerConnection.h>

Public Attributes

Public Attributes Documentation

variable mSystemControllerIP

std::string mSystemControllerIP;

variable mSystemControllerPort

uint16_t mSystemControllerPort;

variable mSystemControllerPostfix

std::string mSystemControllerPostfix;

variable mPSK

std::string mPSK;

variable mUUID

std::string mUUID;

variable mType

ComponentType mType;

variable mName

std::string mName;

variable mMyIP

std::string mMyIP;

variable mConnectTimeout

std::chrono::milliseconds mConnectTimeout {
            3000};

variable mEnableHTTPS

bool mEnableHTTPS;

variable mInsecureHTTPS

bool mInsecureHTTPS;

variable mCustomCaCertFile

std::string mCustomCaCertFile;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.29 - TimeCommon::TAIStatus

TimeCommon::TAIStatus

Public Attributes

Public Attributes Documentation

variable mStratum

StratumLevel mStratum = StratumLevel::UnknownStratum;

variable mHasLock

bool mHasLock = false;

variable mTimeDiffS

double mTimeDiffS = 0.0;

Updated on 2024-04-08 at 15:15:27 +0200

4.1.30 - TimeCommon::TimeStructure

TimeCommon::TimeStructure

Public Attributes

Public Attributes Documentation

variable t1

uint64_t t1 = 0;

variable t2

uint64_t t2 = 0;

variable t3

uint64_t t3 = 0;

variable t4

uint64_t t4 = 0;

variable token

uint64_t token = 0;

variable dummy1

uint64_t dummy1 = 0;

variable dummy2

uint64_t dummy2 = 0;

variable dummy3

uint64_t dummy3 = 0;

Updated on 2024-04-08 at 15:15:27 +0200

4.2 - Files

Files

• dir build
  • dir install
    • dir include
      • file include/AclLog.h
      • file include/AlignedFrame.h
      • file include/Base64.h
      • file include/ControlDataCommon.h
      • file include/ControlDataSender.h
      • file include/DeviceMemory.h
      • file include/IControlDataReceiver.h
      • file include/IMediaStreamer.h
      • file include/ISystemControllerInterface.h
      • file include/IngestApplication.h
      • file include/IngestUtils.h
      • file include/MediaReceiver.h
      • file include/PixelFormat.h
      • file include/SystemControllerConnection.h
      • file include/TimeCommon.h
      • file include/UUIDUtils.h

Updated on 2024-04-08 at 15:15:27 +0200

4.2.1 - include/AclLog.h

include/AclLog.h

Namespaces

Classes

Source code

// Copyright (c) 2022, Edgeware AB. All rights reserved.

#pragma once

#include <filesystem>
#include <string>

#include <spdlog/details/log_msg.h>
#include <spdlog/fmt/fmt.h>
#include <spdlog/formatter.h>
#include <spdlog/pattern_formatter.h>
#include <sys/prctl.h>

namespace AclLog {

enum class Level {
    kTrace,    // Detailed diagnostics (for development only)
    kDebug,    // Messages intended for debugging only
    kInfo,     // Messages about normal behavior (default log level)
    kWarning,  // Warnings (functionality intact)
    kError,    // Recoverable errors (functionality impaired)
    kCritical, // Unrecoverable errors (application must stop)
    kOff       // Turns off all logging
};

void init(const std::string& name);

void initCommandLog(const std::string& name);

void setLevel(Level level);

void logCommand(const std::string& origin, const std::string& command);

AclLog::Level getLogLevel();

size_t getMaxFileSize();

size_t getMaxLogRotations();

std::filesystem::path getLogFileFullPath(const std::string& name);

class CommandLogFormatter : public spdlog::formatter {
public:
    CommandLogFormatter();
    void format(const spdlog::details::log_msg& msg, spdlog::memory_buf_t& dest) override;
    [[nodiscard]] std::unique_ptr<spdlog::formatter> clone() const override;

private:
    std::unique_ptr<spdlog::pattern_formatter> mCommandLogPattern;
};

inline std::string getThreadName() {
    static const size_t RECOMMENDED_BUFFER_SIZE = 20;
    static thread_local std::string name;

    if (name.empty()) {
        char buffer[RECOMMENDED_BUFFER_SIZE];
        int retval = prctl(PR_GET_NAME, buffer);
        if (retval == -1) {
            throw spdlog::spdlog_ex("Failed to get thread name: ", errno);
        }
        name = std::string(buffer);
    }
    return name;
}

class ThreadNameFormatterFlag : public spdlog::custom_flag_formatter {
public:
    void format(const spdlog::details::log_msg&, const std::tm&, spdlog::memory_buf_t& dest) override {
        std::string threadName = getThreadName();
        dest.append(threadName.data(), threadName.data() + threadName.size());
    }

    [[nodiscard]] std::unique_ptr<custom_flag_formatter> clone() const override {
        return spdlog::details::make_unique<ThreadNameFormatterFlag>();
    }
};

class FileLocationFormatterFlag : public spdlog::custom_flag_formatter {
public:
    void format(const spdlog::details::log_msg& msg, const std::tm&, spdlog::memory_buf_t& dest) override {
        if (!msg.source.empty()) {
            using namespace spdlog::details;

            dest.push_back('[');
            const char* filename = short_filename_formatter<null_scoped_padder>::basename(msg.source.filename);
            fmt_helper::append_string_view(filename, dest);
            dest.push_back(':');
            fmt_helper::append_int(msg.source.line, dest);
            dest.push_back(']');
        }
    }

    [[nodiscard]] std::unique_ptr<custom_flag_formatter> clone() const override {
        return spdlog::details::make_unique<FileLocationFormatterFlag>();
    }
};

} // namespace AclLog

Updated on 2024-04-08 at 15:15:27 +0200

4.2.2 - include/AlignedFrame.h

include/AlignedFrame.h

Classes

Types

Functions

Types Documentation

using AlignedAudioFramePtr

using AlignedAudioFramePtr =  std::shared_ptr<AlignedAudioFrame>;

using AlignedAudioFrameConstPtr

using AlignedAudioFrameConstPtr =  std::shared_ptr<const AlignedAudioFrame>;

using AlignedFramePtr

using AlignedFramePtr =  std::shared_ptr<AlignedFrame>;

Functions Documentation

function createAlignedAudioFrame

AlignedAudioFramePtr createAlignedAudioFrame(
    uint8_t numberOfChannels,
    uint32_t numberOfSamplesPerChannel
)