Reporting an update: the design above appears to work as expected, only 17 microseconds apart
Here is the prototype built to the spec above:
![Image]()
Here is the code snippet for the test:And here is the resulting timestamps, only 17 microseconds difference:NOTE: the lux difference is only because the cameras are not both looking at the same object (yet) I have some physical alignment to do still.
Here is the prototype built to the spec above:
Here is the code snippet for the test:
Code:
try: with Picamera2(cam1) as picam1, Picamera2(cam2) as picam2: picam1 = self.configure_picam2(picam1) picam2 = self.configure_picam2(picam2) picam1.start() picam2.start() while True: frame1 = picam1.capture_array() frame2 = picam2.capture_array() lux1 = picam1.capture_metadata()['Lux'] lux2 = picam2.capture_metadata()['Lux'] ts1 = picam1.capture_metadata()['SensorTimestamp'] timestamp1 = self.printable_timestamp(ts1) ts2 = picam2.capture_metadata()['SensorTimestamp'] timestamp2 = self.printable_timestamp(ts2) print(f'\rLux1: {lux1}, Lux2: {lux2}. Cam1 ts: {timestamp1} Cam2 ts: {timestamp2}' , end='') cv2.imshow('Master/Bottom', frame1) cv2.imshow('Slave/Top', frame2) if cv2.waitKey(1) & 0xFF == ord('q'): break except Exception as e:Code:
Lux1: 230.70010375976562, Lux2: 102.81550598144531. Cam1 ts: 2024-02-22 06:34:56.445845 Cam2 ts: 2024-02-22 06:34:56.445862Statistics: Posted by ddbaron — Thu Feb 22, 2024 12:47 pm