BBot V3.1 camera configuration
Prerequisites:
Lens has to be mounted onto the camera.
Camera has to be hooked up with power and the 10GigE to either a laptop or the jetson.
The flash should be mounted with the camera. The flash needs to circumference the lens, while ensure that no stray lights from the flash will enter the optics
The flash should be powered and hooked up with the cameras I/O pins for control. Decision of opto-coupled or not depends on the delay and implementation.
An Xrite/calibrite color checkerboard, placed 2.5 meters away from the lens.
Procedure
Open the aperture fully to minimize the depth of field. Then adjust the focus with great detail until the color checker is exactly in focus. This is easiest in video mode with live feed while zooming in ~400-800% on the distance scales on the color checker.
Set up I/O triggering signals between camera and flash. First, determine the lag from trigger signal until peak in brightness from the flash. Then determine if such a delay is consistent, and whether the trigger signal can be sent ahead of exposure from within the camera. Also measure the flash duration and how the flash intensity varies over time. Be sure to save these plots somewhere.
Assuming a shutter speed of 1/2000 sec, determine how the exposure window is best placed/timed in accordance with the flash duration. Using a second I/O pin and an oscilloscope, verify that the exposure window from the camera matches well with the intensity of the flash.
Set the camera up for fixed exposure settings. This includes:
Exposure speed = 500 microseconds
Gain = 0 dB
Disable all data manipulation features of the camera such as:
White balance = Off
Channel level = 1.0 for R G and B
Gamma correction: 1.0
Set up file formats to raw:
Bayer format with highest available bit-precision. It’s not really important whether packed or non-packed type is used.
Lower the aperture (and don’t touche the focus!) until a good exposure is achieved at the 2.5 meter distance. There should be no overexposed regions, except for maybe the brightest white squares. Take note of this aperture.
Capture sample images of the color checker and save them in 16-bit png-files. They should be and appear greyscale, but show a typical bayerpattern when zoomed in to colored regions in the image.
Finally, vary the distance between the camera and the color checker to evaluate/verify the expected depth of field. This can be done at 1.0, 1.25, 1.5, 1.75, 2, 2.25, 2.37, 2.50, 2.63, 2.75, 3.0 meters distance. Be sure to note the aperture as well as distance to the colorchecker for each raw image that is saved.
To test the clarity/quality of the lens, readjust the color checker distance to 2.5 meters. Then vary the spatial position of the color checker to investigate the level of contrast loss and vignetting we get from non-centered content.