This page provides detailed information related to Design Module 13 - Xilinx Low Latency HLG SDI Audio Video Capture and Display with PL DDR.

Table of Contents

1 Overview

The primary goal of this Design is to demonstrate the capabilities of VCU hard block present in Zynq UltraScale+ EV devices. The TRD will serve as a platform to tune the performance parameters of VCU and arrive at optimal configurations for encoder and decoder blocks. It has also added an initial support of 8-channels audio.

This module enables the capture of the Hybrid Log Gamma(HLG) video from an SDI-Rx subsystem implemented in the PL. The Hybrid Log Gamma(HLG) video can be displayed through the SDI-Tx subsystem implemented in the PL. The module can stream-out and stream-in live captured video frames through an Ethernet interface. This module supports single-stream for XV20 pixel format. In this design, PL_DDR is used for decoding and PS_DDR for encoding so that DDR bandwidth would be enough to support high bandwidth VCU applications requiring simultaneous encoder and decoder operations and transcoding at 4k@60 FPS.

The VCU encoder and decoder operate in slice mode. An input frame is divided into multiple slices (8 or 16) horizontally. The encoder generates a slice_done interrupt at every end of the slice. Generated NAL unit data can be passed to a downstream element immediately without waiting for the frame_done interrupt. The VCU decoder also starts processing data as soon as one slice of data is ready in its circular buffer instead of waiting for complete frame data. The Sync IP does an AXI transaction-level tracking so that the producer and consumer can be synchronized at the granularity of AXI transactions instead of granularity at the video buffer level. Sync IP is responsible for synchronizing buffers between Capture DMA and VCU encoder as both works on the same buffer.

The capture element (FB write DMA) writes video buffers in raster-scan order. SyncIP monitors the buffer level while the capture element is writing into DRAM and allows the encoder to read input buffer data if the requested data is already written by DMA, otherwise it blocks the encoder until DMA completes its writes. On the decoder side, the VCU decoder writes decoded video buffer data into DRAM in block-raster scan order and displays reads data in raster-scan order. To avoid display under-run problems, the software ensures a phase difference of "~frame_period/2", so that decoder is ahead compared to display.

This module supports the Encoding-Decoding and Transmission of Hybrid Log Gamma (HLG) video along with backward compatible Standard Dynamic Range (SDR) for SDI. It provides the ability to encode a wide dynamic range, while still being compatible with the existing transmission standards in the standard dynamic range (SDR) region. This HLG format encodes the HDR and SDR information in single signal enabling HDR-compatible TVs to display an enhanced image. Unlike HDR it does not have any metadata, rather it will use the Alternative transfer characteristics (ATC) and Supplemental Enhanced Information (SEI) in the Video Usability Information (VUI) to add extra encoding details.

From VCU point of view, there are two "types" of HLG, which you can enable:

1. There is a HLG-SDR Backwards Compatible Mode, which uses the BT2020 value in the SPS VUI parameters instead of the HLG transfer characteristics. Then the VCU encoder will insert an 'Alternative Transfer Characteristics' (ATC) SEI with the HLG value. See the below video frame snapshot captured in the stream-eye:

Open

Open

0x93 - Payload Type (147 == ATC)
0x01 - Payload Size (1 byte)
0x12 - 18 (HLG EOTF value)
0x80 - payload bits ending

2. There is a HLG only mode. This directly uses the HLG value in the SPS VUI parameters. See below frame snapshot captured in stream-eye:

This design supports the following video interfaces:
Sources:

• SDI-Rx capture pipeline implemented in the PL.

• File source (SD card, USB storage, SATA hard disk).

• Stream-In from network or internet.

Sinks:

• SDI-Tx display pipeline implemented in the PL.

VCU Codec:

• Video Encode/Decode capability using VCU hard block in PL 

  • AVC/HEVC encoding.

  • Encoder/decoder parameter configuration.

Streaming Interfaces:

• 1G Ethernet PS GEM 

Video format:

• XV20

Audio Configuration:

• Codec: Opus

• Format: S24_32LE

• Channels: 2, 8

• Sampling rate: 48 kHz

Supported Resolution

The table below provides the supported resolution from GUI and command-line app in this design.

√ - Supported
x – Not supported

The below table gives information about the features supported in this design. 

The below figure shows the HLG SDI Video Capture and HLG SDI Display with Audio design hardware block diagram.

The below figure shows the HLG SDI Video Capture and HLG SDI Display with Audio design software block diagram.

1.1 Board Setup

Refer to the below link for board setup

• Zynq UltraScale+ MPSoC VCU TRD 2022.1 Board Setup

1.2 Run Flow

The TRD package is released with the source code, Vivado project, PetaLinux BSP, and SD card image that enables the user to run the demonstration. It also includes the binaries necessary to configure and boot the ZCU106 board. Prior to running the steps mentioned in this wiki page, download the TRD package and extract its contents to a directory referred to as TRD_HOME - which is the home directory.

Refer to the below link to download all TRD contents.

• Refer to Section 4.1 : Download the TRD of Zynq UltraScale+ MPSoC VCU TRD 2022.1 wiki page to download all TRD contents.

TRD package contents specific to HLG SDI Video Capture and HLG SDI Display with Audio design are placed in the following directory structure. The user needs to copy all the files from the $TRD_HOME/images/vcu_llp2_hlg_sdi to FAT32 formatted SD card directory.

TRD package contents specific to Xilinx Low Latency PL DDR HLG SDI Audio-Video design is placed in the following directory structure.

Configuration files (input.cfg) for various Resolutions are placed in the following directory structure in /media/card.

1.2.1 GStreamer Application (vcu_gst_app)

The vcu_gst_app is a command-line multi-threaded Linux application. The command-line application requires an input configuration file (input.cfg) to be provided in the plain text.

Execution of the application is shown below:

Example:

4kp60 HEVC_HIGH Display Pipeline execution

4kp60 HEVC_HIGH Record Pipeline execution

4kp60 HEVC_HIGH Stream-out Pipeline execution

4kp60 HEVC_HIGH Stream-in Pipeline execution

LLP2 4kp60 HEVC_HIGH Display Pipeline execution

LLP2 4kp60 HEVC_HIGH Stream-out Pipeline execution

LLP2 4kp60 HEVC_HIGH Stream-in Pipeline execution

To measure the latency of the pipeline, run the below command. The latency data is huge, so dump it to a file.

Refer to the below link for detailed run flow steps

• Zynq UltraScale+ MPSoC VCU TRD 2022.1 - Run Flow

1.3 Build Flow

Refer to the below link for build flow

• Zynq UltraScale+ MPSoC VCU TRD 2022.1 - Build Flow

2 Other Information

2.1 Known Issues

• For PetaLinux related known issues please refer to: PetaLinux 2022.1 - Product Update Release Notes and Known Issues

• For VCU related known issues please refer to Answer Record 76600: LogiCORE H.264/H.265 Video Codec Unit (VCU) - Release Notes and Known Issues and Xilinx Zynq UltraScale+ MPSoC Video Codec Unit.

2.2 Limitations

• For PetaLinux related limitations please refer to: PetaLinux 2022.1 - Product Update Release Notes and Known Issues

• For VCU related limitations please refer to Answer Record 76600: LogiCORE H.264/H.265 Video Codec Unit (VCU) - Release Notes and Known Issues , Xilinx Zynq UltraScale+ MPSoC Video Codec Unit and PG252.

2.3 Optimum VCU Encoder parameters for use-cases:

Video streaming:

• Video streaming use-case requires a very stable bitrate graph for all pictures

• It is good to avoid periodic large Intra pictures during the encoding session

• Low-latency rate control (hardware RC) is the preferred control-rate for video streaming, it tries to maintain equal amount frame sizes for all pictures.

• Good to avoid periodic Intra frames instead use low-delay-p (IPPPPP…)

• VBR is not a preferred mode of streaming

Performance: AVC Encoder settings:

• It is preferred to use 8 or higher slices for better AVC encoder performance

• AVC standard does not support Tile mode processing which results in the processing of MB rows sequentially for entropy coding

Quality: Low bitrate AVC encoding:

• Enable profile=high and use qp-mode=auto for low-bitrate encoding use-cases

• The high profile enables 8x8 transform which results in better video quality at low bitrate

2.4 Audio-Video Synchronization

Clocks and synchronization in GStreamer

When playing complex media, each sound and video sample must be played in a specific order at a specific time. For this purpose, GStreamer provides a synchronization mechanism.

GStreamer provides support for the following use cases:

• Non-live sources with access faster than playback rate. This is the case where one is reading media from a file and playing it back in a synchronized fashion. In this case, multiple streams need to be synchronized, like audio, video and subtitles.

• Capture and synchronized muxing/mixing of media from multiple live sources. This is a typical use case where you record audio and video from a microphone/camera and mux it into a file for storage.

• Streaming from (slow) network streams with buffering. This is the typical web streaming case where you access content from a streaming server using HTTP.

• Capture from live source and playback with configurable latency. This is used, for example, when capturing from a camera, applying an effect, and displaying the result. It is also used when streaming low latency content over a network with UDP.

• Simultaneous live capture and playback from prerecorded content. This is used in audio recording cases where you play a previously recorded audio and record new samples, the purpose is to have the new audio perfectly in sync with the previously recorded data.

GStreamer uses a GstClock object, buffer timestamps and a SEGMENT event to synchronize streams in a pipeline as we will see in the next sections.

Clock running-time

In a typical computer, there are many sources that can be used as a time source, e.g., the system time, soundcards, CPU performance counters, etc. For this reason, GStreamer has many GstClock implementations available. Note that clock time doesn't have to start from 0 or any other known value. Some clocks start counting from particular start date, others from the last reboot, etc.

A GstClock returns the absolute-time according to that clock with gst_clock_get_time (). The absolute-time (or clock time) of a clock is monotonically increasing.

A running-time is the difference between a previous snapshot of the absolute-time called the base-time and any other absolute-time.

running-time = absolute-time - base-time

A GStreamer GstPipeline object maintains a GstClock object and a base-time when it goes to the PLAYING state. The pipeline gives a handle to the selected GstClock to each element in the pipeline along with selected base-time. The pipeline will select a base-time in such a way that the running-time reflects the total time spent in the PLAYING state. As a result, when the pipeline is PAUSED, the running-time stands still.

Because all objects in the pipeline have the same clock and base-time, they can thus all calculate the running-time according to the pipeline clock.

Buffer running-time

To calculate a buffer running-time, we need a buffer timestamp and the SEGMENT event that preceded the buffer. First we can convert the SEGMENT event into a GstSegment object and then we can use the gst_segment_to_running_time () function to perform the calculation of the buffer running-time.

Synchronization is now a matter of making sure that a buffer with a certain running-time is played when the clock reaches the same running-time. Usually, this task is performed by sink elements. These elements also have to take into account the configured pipeline's latency and add it to the buffer running-time before synchronizing to the pipeline clock.

Non-live sources timestamp buffers with a running-time starting from 0. After a flushing seek, they will produce buffers again from a running-time of 0.

Live sources need to timestamp buffers with a running-time matching the pipeline running-time when the first byte of the buffer was captured.

Buffer stream-time

The buffer stream-time, also known as the position in the stream, is a value between 0 and the total duration of the media and it's calculated from the buffer timestamps and the preceding SEGMENT event.

The stream-time is used in:

• Report the current position in the stream with the POSITION query.

• The position used in the seek events and queries.

• The position used to synchronize controlled values.

The stream-time is never used to synchronize streams, this is only done with the running-time.

Time overview

Here is an overview of the various timelines used in GStreamer.

The image below represents the different times in the pipeline when playing a 100ms sample and repeating the part between 50ms and 100ms.

You can see how the running-time of a buffer always increments monotonically along with the clock-time. Buffers are played when their running-time is equal to the clock-time - base-time. The stream-time represents the position in the stream and jumps backwards when repeating.

Clock providers

A clock provider is an element in the pipeline that can provide a GstClock object. The clock object needs to report an absolute-time that is monotonically increasing when the element is in the PLAYING state. It is allowed to pause the clock while the element is PAUSED.

Clock providers exist because they play back media at some rate, and this rate is not necessarily the same as the system clock rate. For example, a sound card may play back at 44.1 kHz, but that doesn't mean that after exactly 1 second according to the system clock, the sound card has played back 44100 samples. This is only true by approximation. In fact, the audio device has an internal clock based on the number of samples played that we can expose.

If an element with an internal clock needs to synchronize, it needs to estimate when a time according to the pipeline clock will take place according to the internal clock. To estimate this, it needs to slave its clock to the pipeline clock.

If the pipeline clock is exactly the internal clock of an element, the element can skip the slaving step and directly use the pipeline clock to schedule playback. This can be both faster and more accurate. Therefore, generally, elements with an internal clock like audio input or output devices will be a clock provider for the pipeline.

When the pipeline goes to the PLAYING state, it will go over all elements in the pipeline from sink to source and ask each element if they can provide a clock. The last element that can provide a clock will be used as the clock provider in the pipeline. This algorithm prefers a clock from an audio sink in a typical playback pipeline and a clock from source elements in a typical capture pipeline.

There exist some bus messages to let you know about the clock and clock providers in the pipeline. You can see what clock is selected in the pipeline by looking at the NEW_CLOCK message on the bus. When a clock provider is removed from the pipeline, a CLOCK_LOST message is posted and the application should go to PAUSED and back to PLAYING to select a new clock.

For more detail please refer to: https://gstreamer.freedesktop.org/documentation/application-development/advanced/clocks.html?gi-language=c

3 Appendix A - Input Configuration File (input.cfg)

The example configuration files are stored at /media/card/config/ folder.