Quick Start Guide: Prime Color Setup
This page provides instructions on how to set up, configure, and use the Prime Color video camera.
The Prime Color is a full-color video camera that is capable of recording synchronized high-speed and videos. It can also be hooked up to a mocap system and used as a reference camera. The camera enables recording of high frame rate videos (up to 500 FPS at 480p) with resolutions up to 1080p (at 250 FPS) by performing onboard compression (H.264) of captured frames. It connects to the camera network and receives power by a standard PoE connection.
When capturing high-speed videos, the time-length of camera exposures are very short, and thus, providing sufficient lighting becomes critical for obtaining clear images. The eStrobe is designed to optimally brighten the image taken by Prime Color camera by precisely synchronizing the illuminations of the eStrobe LEDs to each camera exposure. This allows the LEDs to illuminate at a right timing, producing the most efficient and powerful lighting for the high-speed video capture. Also, the eStrobe emits white light only, and it will not interfere with the tracking within the IR spectrum.
Required PC specifications may vary depending on the size of the camera system. Generally, you will be required to use the recommended specs with a system with more than 24 cameras.
For using Prime Color cameras, it requires the computer to be equipped with a dedicated graphics card that has a performance of GTX 1050, or better, with the latest driver that supports OpenGL version 4.0 or higher.
Since each color camera can upload a large amount of data over the network, the size of the recorded Take (TAK) can get pretty large even with a short recording. For example, if a 10-second take was recorded with a total data throughput of 1-GBps, the resulting TAK file will be 10-GB, and it can quickly fill up the storage device. Please make sure there is enough capacity available on the disk drive. If you are exporting out the recorded data onto video files after they are captured, re-encoding the videos will help with reducing the files magnitudes smaller. See: Re-encoding
Since Prime Color cameras can output a large amount of data to the RAM memory quickly, it is also important that the write-out speed to the storage is also fast enough. If the write-out speed to secondary drive isn't fast enough, the occupied memory in RAM storage may gradually increase to its maximum. For recording with just a one or two Prime Color cameras, standard SSD drive will do its job. However, when using multiple Prime Color cameras, it is recommended to use a fast storage drive (e.g. M.2 SSD) that can quickly write out the recorded capture that from the RAM.
When running two or more Prime Color cameras, the computer must have a 10-gigabit network adapter in order to successfully receive all of the data outputted from the camera system. Please see Load Balancing section for more information.
Different types of lenses can be equipped on a Prime Color camera as long as the lens mount is compatible, however, for Prime Color cameras, we suggest using C-mount lenses to fully utilize the imager. Prime Color cameras with C-mount can be equipped with either the 12mm F#1.8 lenses or the 6.8mm F#1.6 lenses. The 12mm lens is zoomed in more and is more suitable for capturing at long ranges. On the other hand, the 6.8mm lens has a larger field of view and is more suitable for capturing a wide area. Both lenses have adjustable f-stop and focus settings, which can be optimized for different capture environments and applications.
- F-Stop: Set the f-stop to a low value to make the aperture size bigger. This will allow in more light onto the imager, improving the image quality. However, this may also decrease the camera's depth of field, requiring the lens to be focused specifically on the target capture area.
- Focus: For best image quality, make sure the lenses are focused on the target tracking area.
Before going into details of setting up a system with Prime Color cameras, it is important to go over the data bandwidth availability within the camera network. At its maximum bit-rate setting for capturing the best quality image, one Prime Color camera can transmit data at a rate of up to ~100 Megabytes-per-second (MBps), or ~800 Megabits-per-second (Mbps). For a comparison, a tracking camera in Object Mode outputs data at a rate less than 1MBps, which is several magnitudes smaller than the output from a Prime Color camera. A standard network switch (1 Gb switch) and network card only support network traffic of up to 1000 Mbps (or 1 Gbps). When Prime Color camera(s) are used, they can take up a large portion, or all, of the available bandwidth, and for this reason, extra attention to bandwidth use will be needed when first setting up the system.
When there is not enough available bandwidth, captured 2D frames may drop out due to the data bottleneck. Thus, it is important to take the bandwidth consumption into account and make sure an appropriate set of network switches (PoE and Uplink), Ethernet cables, and a network card is used. If a 1-Gb network/uplink switch is used, then only one Prime Color camera can be used at its maximum bit-rate setting. If two or more Prime Color cameras need to be used, then either a 10-Gb network setup will be required OR the bit-rate setting will need to be turned down. A lower bit-rate will further compress the image with a tradeoff on the image quality, which may or may not be acceptable depending on the capture application.
Each Prime Color camera must be uplinked and powered through a standard PoE connection that can provide at least 15.4 watts to each port simultaneously.
One or Two Color Cameras
Multiple Color Cameras
Prime Color cameras connect to the camera system just like other Prime series camera models. Simply plug the camera onto a PoE switch that has enough available bandwidth and it will be powered and synchronized along with other tracking cameras. When you have two color cameras, they will need to be distributed evenly onto different PoE switches so that the data load is balanced out.
Click image to enlarge.
When using multiple Prime Color cameras, we recommend connecting the color cameras directly into the 10-gigabit aggregation (uplink) switch, because such setup is best for preventing bandwidth bottleneck. A PoE injector will be required if the uplink switch does not provide PoE. This allows the data to travel directly onto the uplink switch and to the host computer through the 10-gigabit network interface. This will also separate the color cameras from the tracking cameras.
Click image to enlarge.
The eStrobe synchronizes with Prime Color cameras through RCA cable connection. It receives exposure signals from the cameras and synchronizes its illuminations correspondingly. Depending on the frame rate of the camera system, the eStrobe will vary its illumination frequency, and it will also vary the percent duty cycle depending on the exposure length. Multiple eStrobes can be daisy-chained in series by relaying the sync signal from the output port to the input port of another as shown in the diagram.
eStrobe with Single Prime Color Camera
eStrobe with Multiple Prime Color Cameras
Click image to enlarge.
Click image to enlarge.
When capturing without eStrobes, the camera entirely relies on the ambient lighting to capture the image, and the brightness of the captured frames may vary depending on which type of light source is used. In general, when capturing without an eStrobe, we recommend setting the camera at a lower framerate (30~120 FPS) and increasing the camera exposure to allow for longer exposure time so that the imager can take in more light.
When capturing indoors without the eStrobe, you will be relying on the room lighting for brightening up the volume. Here, it is important to note that every type of artificial light source illuminates, or flickers, at a certain frequency (e.g. fluorescent light bulbs typically flicker at 120Hz). This is usually fast enough so that the flickering is not noticeable to human eyes, however, with high-speed cameras, the flickering may become apparent.
When Prime Color captures at a frame rate higher than the ambient illumination frequency, you will start noticing brightness changes between consecutive frames. This happens because, with mismatching frequencies, the cameras are exposing at different points of the illumination phase. For example, if you capture at 240FPS with 120Hz light bulbs lighting up the volume, brightness of captured images may be different in even and odd numbered frames throughout the capture. Please take this into consideration and provide appropriate lighting as needed.
Incandescent light flickering. The video was captured at (121 FPS).
When capturing outdoors using Prime Color cameras, sunlight will typically provide enough ambient lighting. Unlike light bulbs, sunlight is emitted continuously, so there is no need to worry about the illumination frequency. Furthermore, the sun is bright enough and you should be able to capture high-quality images by adjusting only the f-stop (aperture size) and the exposure values.
Now that you have set up a camera system with Prime Color, all of the connected cameras should be listed under the Devices pane. At this point, you would want to launch Motive and check the following items to make sure your system is operating properly.
- 2D Frame Delivery: There should be no dropped 2D frames. You can monitor this under the Log pane or from the Devices pane. If frame drops are reported continuously, you can lower down the bit-rate setting or revisit the network configuration and make sure the data loads are balanced out. For more information, Data Bandwidth section of this page.
- CPU Usage: Open the windows task manager and check the CPU processing load. If only one of the CPU core is fully occupied, the CPU is not fast enough to process data from the color camera. In this case, you will want to use a faster CPU or lower down the bit-rate setting.
- RAM Usage: Open the windows task manager and check the memory usage. If the RAM usage slowly creeps up to the maximum memory while recording a take, it means the disk driver is not fast enough to write out the color video from RAM. You will have to reduce the bit-rate setting or use a faster disk drive (e.g. M.2 SSD).
- Hard Drive Space: Make sure there is enough memory capacity available on the computer. Take files (TAK) with color camera data can be quite large, and it could quickly fill up the memory, especially, when recording lightly-compress video from multiple color cameras.
When you launch Motive, connected Prime Color cameras will be shown in Motive, and you will be able to configure the settings as you would do for other tracking cameras. Open up the Devices pane and the Properties pane, and select a Prime Color camera(s). On the Properties pane, key properties that are specific to the selected color cameras will be listed. Optimizing these settings are important in order to obtain best quality images without overflooding the network bandwidth. The key settings for the color cameras are image resolution, gamma correction, as well as compression mode and bit-rate settings, which will be covered in the following sections.
Default: 1920, 1080
This property sets the resolution of the images that are captured by selected cameras. Since the amount of data increases with higher resolution, depending on which resolution is selected, the maximum allowable frame rate will vary. Below is the maximum allowed frame rates for each respective resolution setting.
Default: Constant Bit Rate.
This property determines how much the captured images will be compressed. The Constant Bit-Rate mode is used by default and recommended because it is easier to control the data transfer rate and efficiently utilize the available network bandwidth.
In the Constant Bit-Rate mode, Prime Color cameras vary the degree of image compression to match the data transmission rate given under the Bit Rate settings. At a higher bit-rate setting, the captured image will be compressed less. At a lower bit-rate setting, the captured image will be compressed more to meet the given data transfer rate, but compression artifacts may be introduced if it is set too low.
Variable Bit-Rate setting is also available for keeping the amount of the compression constant and allowing the data transfer rate to vary. This mode can be beneficial when capturing images with objects that have detailed textures because it keeps the amount of compression same on all frames. However, this may introduce dropped frames whenever the camera tries to compress highly detailed images because it will increase the data transfer rate; which may overflow the network bandwidth as a result. For this reason, we recommend using the Constant Bit-Rate setting in most applications.
Available only while using Constant Bit-rate Mode
Bit-rate setting determines the transmission rate outputted from the selected color camera. The value given under this setting is measured in percentage (100%) of the maximum data transmission speed, and each color camera can output up to ~100 MBps. In other words, the configured value will indirectly represent the transmission rate in Megabytes per second (MBps). At bit-rate setting of 100, the camera will capture the best quality image, however, it could overload the network if there is not enough bandwidth to handle the transmitted data.
Since the bit-rate controls the amount of data outputted from each color camera, this is one of the most important settings when properly configuring the system. If your system is experiencing 2D frame drops, it means one of the system requirements is not met; either network bandwidth, CPU processing, or RAM/disk memory. In such cases, you could decrease the bit-rate setting and reduce the amount of data output from the color cameras.
The image quality will increase at a higher bit-rate setting because it records a larger amount of data, but this will result in large file sizes and possible frame drops due to data bandwidth bottleneck. Often, the desired result is different depending on the capture application and what it is used for. The below graph illustrates how the image quality varies depending on the camera framerate and bit-rate settings.
Default : 24
Gamma correction is a non-linear amplification of the output image. The gamma setting will adjust the brightness of dark pixels, midtone pixels, and bright pixels differently, affecting both brightness and contrast of the image. Depending on the capture environment, especially with a dark background, you may need to adjust the gamma setting to get best quality images.
If you are using the eStrobes to light up the capture volume, the LED setting must be enabled on the Prime Color cameras which the eStrobes connect to. When this setting is enabled, the Prime Color camera will start outputting the signals from its RCA sync output port, allowing the eStrobes to receive this signal and illuminate the LEDs.
In order to calibrate the color camera into the 3D capture volume, the Prime Color camera must be equipped with an IR filter switcher. Prime Color cameras without IR filter switcher cannot be calibrated, and can only be used as a reference camera to monitor the reference views in the 2D Camera View pane or in the Cameras viewport.
The Prime Color FS is equipped with a filter switcher that allows the cameras to detect in IR spectrum. The Prime Color FS can be calibrated into the 3D capture volume using an active calibration wand with the IR LEDs. Once calibrated, the color camera will be placed within the 3D viewport along with other tracking cameras, and 3D assets (Marker Sets, Rigid Body, Skeletons, cameras) can be overlaid as shown in the image.
To calibrate the camera, switch the Prime Color FS to the Object mode in the Camera Preview pane. This will switch the Color camera to detect in the IR spectrum, and then use the active wand to follow the standard Calibration process. Once the calibration is finished, you can switch the camera back to the Color Video Mode.
Reference video zoomed into a Rigid Body.
Once you have set up the system and configured the cameras correctly, Motive is now ready to capture Takes. Recorded TAK files will contain color video along with the tracking data, and you can play them back in Motive. Also, the color reference video can be exported out from the TAK.
Once the camera is set up, you can start recording from Motive. Captured frames will be stored within the TAK file and you can access them again in Edit mode. Please note that capture files with Prime Color video images will be much larger in file size.
Once the color videos have been saved onto TAK files, the captured reference videos can be exported into AVI files using either H.264 or MJPEG compression format. The H.264 format will allow faster export of the recorded videos and is recommended. Video for the current TAK can be exported by clicking File tab -> Export Video option in Motive, or you can also export directly from the Data pane by right-clicking on the Take(s) and clicking Export Video from the context menu. The following export dialogue window will open and you will be able to configure the export settings before outputting the files:
When this is set to Drop Frames, Motive will remove any dropped frames in the color video upon export. Please note that any dropped frames will be completely removed in this case, and thus, the exact frames in the exported file may not match the frames in the corresponding Motive recording. If needed, you can set this export option to Black Frame to insert black, or blank, frames in place of the dropped frames in the exported video.
If there are multiple TAK files containing reference video recordings, you can export the videos all at once in the Data pane or through the Motive Batch Processor. When exporting directly from the Data pane, simply CTRL-select multiple TAK files together, right-click to bring up the context menu, and click Export Video. When using the batch processor (NMotive), the VideoExporter class can be used to export videos from loaded TAK files.
The size of the exported video file can be re-encoded and compressed down further by additional subsampling. This can be achieved using a third-party video processing software, and doing so can hugely reduce the size of the exported file; almost in orders of two magnitudes. This is supported by most of the high-end video editing software, but Handbrake (https://handbrake.fr/) is a freely available open-source software that is also capable of doing this. Since the exported video file can be large in size, we suggest using one of the third-party software to re-encode the exported video file.
Slow memory write out.