DRIVE OS Version: 6.0.10
Question:
If the following questions are outside NVIDIA’s scope of support (for example, camera sensor vendor specific behavior), please let me know. I need a clear answer based on specifications or implementation details rather than assumptions.
1. Interpretation of SOF and EOF timestamps
In the following forum post:
the following explanation is given:
Exposure Start Timestamp = sofTimestampUs - exposureDurationUs[0]
Can the same interpretation be applied to eofTimestampUs?
In other words, is the following interpretation correct?
・first row exposure start = sofTimestampUs - exposureDurationUs[0]
・last row exposure start = eofTimestampUs - exposureDurationUs[0]
2. Interpretation of rolling shutter timing
If the answer to Question 1 is “Yes”, I would like to confirm whether my understanding of the timing diagram is correct.
Does this indicate that:
・The exposure start time of the first row and the last row are different?
・When the exposure time is 10 ms, the exposure period of the last row of Frame 1 overlaps in time with the exposure period of the first row of Frame 2?
Camera Configuration
Sensor : IMX728
Resolution : 3840 x 2160
Frame Rate : 30 FPS
dwSensorCamera_getImageTimestamps()
sofTimestampUs : 1773907931.235 s
eofTimestampUs : 1773907931.265 s
eofTimestampUs - sofTimestampUs = 30 ms
dwImage_getMetaData()
exposureDurationUs[0] : 10.000 ms (Long)
exposureDurationUs[1] : 10.000 ms (Middle)
exposureDurationUs[2] : 0.055 ms (Short)
Based on the above values, I created the attached timing diagram.
My interpretation is that:
・Is it correct that there is a timing offset between the exposure start times of the first row and the last row?
・With an exposure time of 10 ms, the exposure period of the last row of Frame 1 overlaps with the exposure period of the first row of Frame 2.
Could you please confirm whether this interpretation is correct?
