Balluff MATRIX VISION mvBlueNAOS Technical Manual page 204

196
Figure 1: mvTimestampReset
This can be elegantly used for synchronization purposes by means of wiring an input of all cameras together
and reset all camera timestamps at the beginning by a defined signal edge from the process. From this reset on
all cameras start at zero local time and will increment independently their timestamp so that we achieve a basic
accuracy only limited by drift of the clock main frequency (e.g. a 1 MHz oscillator in the FPGA) over time.
In order to compensate for this drift we can in addition reset the timestamp every second or minute or so and count
the reset pulse itself by a counter in each camera. Assuming this reset pulse is generated by the master camera
itself by means of a timer and output as the hardware reset signal for all cameras, we now can count the reset pulse
with all cameras and put both the count and the reset timestamp as so called chunk data in the images.
We thus have achieved a synchronized timestamp with the precision of the master camera among all connected
cameras.
Settings required are shown using MATRIX VISION's wxPropView (p. 65) tool:
Figure 2: Reset the timestamp every second
MATRIX VISION GmbH