Ping was missed RuntimeError when using RGB calibrated WarpMesh with OAK-D PoE

danl5

I recently tried to get undistortion running on board an OAK-D PoE by explicitly setting the warp mesh source for the RGB camera to the CALIBRATION value: cam_rgb.setMeshSource(dai.CameraProperties.WarpMeshSource.CALIBRATION). When I add that line to my code, my camera usually only runs for less than 5 minutes before I get a "ping was missed" RuntimeError: Communication exception. If I remove the line, things seem stable and run apparently indefinitely. I also tried with another OAK-D PoE and saw the same problem.

Here is what I've tried:

Enabled debug and checked CPU usage (<60% all the way up to the crash, nowhere close to 90%+ as warned against in several other posts).
Upgrade the bootloader: I was on bootloader 0.0.24, but upgraded to 0.0.26 and tried again. Still get the Communication exception.
Upgrade depthai: started with 2.23.0.0, also tried 2.25.0.0. Still get the exception.
Set the watchdog timeout to a max of 4500 by setting the environment variable. Also no improvement.

Configuration Details:

RGB camera at 4K, 12fps encoded with H.265 to keep bandwidth utilization manageable
Stereo depth output at 720p, 12fps with extended disparity and aligned to the RGB camera
IMU Rotation Vector, Linear accel, and calibrated gyroscope sensors enabled at 100Hz

I've attached the code I use that consistently results in the problem. (change the SETWARPMESH = True to False on line 20 to excluding the warp mesh setting.

Any help resolving this would be appreciated!

import datetime
import logging

import depthai as dai

# Queue Names and Config
QUEUE_SIZE_MULTIPLIER = 4
COLOR_QUEUE = 'rgb_video'
COLOR_CONTROL_QUEUE = 'rgb_control'
DEPTH_QUEUE = 'depth'
MONO_LEFT_CONTROL_QUEUE = 'mono_left_control'
MONO_RIGHT_CONTROL_QUEUE = 'mono_right_control'
IMU_QUEUE = 'imu'

# Color Config
COLOR_SIZE = (3840,2160)
COLOR_FPS = 12
COLOR_CODEC = dai.VideoEncoderProperties.Profile.H265_MAIN
COLOR_BITRATE = 20000000 # 20 Mbps
SET_WARP_MESH = True # whether or not to set the warp mesh for onboard undistort

# Stereo Depth Config
MONO_RESOLUTION = dai.MonoCameraProperties.SensorResolution.THE_720_P
MONO_SIZE = (1280,720)
MONO_FPS = 12

# Imu Config
IMU_SENSORS = [dai.IMUSensor.ROTATION_VECTOR,
               dai.IMUSensor.LINEAR_ACCELERATION,
               dai.IMUSensor.GYROSCOPE_CALIBRATED]
IMU_RATE = 100

# Logging Config
DEPTHAI_LOG_LEVEL = dai.LogLevel.DEBUG

logging.basicConfig(
    level="INFO",
    format='%(asctime)s:%(levelname)s:%(name)s:%(message)s',
    datefmt='%Y-%m-%d %H:%M:%S'
)

logger = logging.getLogger(__name__)

def configure_pipeline(pipeline: dai.Pipeline):
    #########################
    # RGB Setup
    #########################
    cam_rgb = pipeline.create(dai.node.Camera)

    # Properties
    cam_rgb.setBoardSocket(dai.CameraBoardSocket.CAM_A) # CAM_A is RGB Camera
    cam_rgb.setSize(COLOR_SIZE)
    cam_rgb.setFps(COLOR_FPS)

    # Define sources and outputs
    control_in = pipeline.createXLinkIn()
    control_in.setMaxDataSize(1)
    control_in.setStreamName(COLOR_CONTROL_QUEUE)
    control_in.out.link(cam_rgb.inputControl)

    video_encoder = pipeline.create(dai.node.VideoEncoder)
    video_encoder.setDefaultProfilePreset(COLOR_FPS, COLOR_CODEC)
    video_encoder.setRateControlMode(dai.VideoEncoderProperties.RateControlMode.CBR)
    video_encoder.setBitrate(COLOR_BITRATE)

    rgb_out = pipeline.createXLinkOut()
    rgb_out.setStreamName(COLOR_QUEUE)
    cam_rgb.video.link(video_encoder.input)
    video_encoder.bitstream.link(rgb_out.input)

    if SET_WARP_MESH:
        logger.info("Configuring RGB camera for on-board undistortion")
        cam_rgb.setMeshSource(dai.CameraProperties.WarpMeshSource.CALIBRATION)
    else:
        logger.info("RGB camera will NOT be configured for on-board undistortion")


    rgb_out.input.setBlocking(False)
    rgb_out.input.setQueueSize(COLOR_FPS * QUEUE_SIZE_MULTIPLIER)

    #########################
    # Stereo Depth Setup
    #########################
    left = pipeline.createMonoCamera()
    right = pipeline.createMonoCamera()

    left.setResolution(MONO_RESOLUTION)
    left.setBoardSocket(dai.CameraBoardSocket.CAM_B) # CAM_B is Left Camera
    left.setFps(MONO_FPS)
    right.setResolution(MONO_RESOLUTION)
    right.setBoardSocket(dai.CameraBoardSocket.CAM_C) # CAM_C is Right Camera
    right.setFps(MONO_FPS)

    left_control_in = pipeline.createXLinkIn()
    left_control_in.setMaxDataSize(1)
    left_control_in.setStreamName(MONO_LEFT_CONTROL_QUEUE)

    right_control_in = pipeline.createXLinkIn()
    right_control_in.setMaxDataSize(1)
    right_control_in.setStreamName(MONO_RIGHT_CONTROL_QUEUE)

    left_control_in.out.link(left.inputControl)
    right_control_in.out.link(right.inputControl)

    stereo = pipeline.createStereoDepth()
    left.out.link(stereo.left)
    right.out.link(stereo.right)

    stereo.setDefaultProfilePreset(dai.node.StereoDepth.PresetMode.HIGH_DENSITY)
    stereo.initialConfig.setMedianFilter(dai.StereoDepthProperties.MedianFilter.KERNEL_3x3)
    stereo.setExtendedDisparity(True)
    stereo.setLeftRightCheck(True)
    stereo.setDepthAlign(dai.CameraBoardSocket.CAM_A)

    stereo.setOutputSize(*MONO_SIZE)

    depth_out = pipeline.createXLinkOut()
    depth_out.setStreamName(DEPTH_QUEUE)
    stereo.depth.link(depth_out.input)
    depth_out.input.setBlocking(False)
    depth_out.input.setQueueSize(MONO_FPS * QUEUE_SIZE_MULTIPLIER)

    #########################
    # IMU Setup
    #########################
    imu = pipeline.create(dai.node.IMU)
    imu_out = pipeline.create(dai.node.XLinkOut)

    imu_out.setStreamName(IMU_QUEUE)

    for which_sensor in IMU_SENSORS:
        imu.enableIMUSensor(which_sensor, IMU_RATE)
    imu.setBatchReportThreshold(20)
    imu.setMaxBatchReports(20)
    imu.out.link(imu_out.input)

    imu_out.input.setBlocking(False)
    imu_out.input.setQueueSize(IMU_RATE * QUEUE_SIZE_MULTIPLIER)


def processing_loop(oak_device: dai.Device):
    rgb_video_queue = oak_device.getOutputQueue(COLOR_QUEUE)
    depth_queue = oak_device.getOutputQueue(DEPTH_QUEUE)
    imu_queue = oak_device.getOutputQueue(IMU_QUEUE)

    last_rgb_sequence_number = -1
    last_depth_sequence_number = -1
    last_imu_sequence_number = -1

    logger.info("Listening for packets...")
    while True:
        oak_device.getQueueEvent([COLOR_QUEUE, DEPTH_QUEUE, IMU_QUEUE],
                                    datetime.timedelta(seconds=2))

        rgb_packet = rgb_video_queue.tryGet()
        depth_packet = depth_queue.tryGet()
        imu_packet = imu_queue.tryGet()

        for queue_name, packet, last_seq_num in [(COLOR_QUEUE, rgb_packet, last_rgb_sequence_number),
                                                 (DEPTH_QUEUE, depth_packet, last_depth_sequence_number),
                                                 (IMU_QUEUE, imu_packet, last_imu_sequence_number)]:
            if packet:
                latest_seq_num = packet.getSequenceNum()
                sequence_step = latest_seq_num - last_seq_num
                if sequence_step > 1 and last_seq_num >= 0:
                    # only consider it a missed sequence if we've received our first packet
                    # (last_packet_sequence number will be -1 if no packet yet received)
                    logger.info("Received non-contiguous sequence numbers in the %s queue with a difference of %d (%d and %d)",
                                queue_name, sequence_step, last_seq_num, latest_seq_num)
                last_seq_num = latest_seq_num


def main():
    # Create pipeline
    pipeline = dai.Pipeline()

    configure_pipeline(pipeline)

    logger.info("Pipline configured, sending to camera...")
    oak_device = dai.Device(pipeline)
    oak_device.setLogLevel(DEPTHAI_LOG_LEVEL)
    oak_device.setLogOutputLevel(DEPTHAI_LOG_LEVEL)

    rgb_control_queue = oak_device.getInputQueue(COLOR_CONTROL_QUEUE)


    ctrl = dai.CameraControl()
    ctrl.setAntiBandingMode(dai.CameraControl.AntiBandingMode.MAINS_60_HZ)
    ctrl.setAutoExposureEnable()
    ctrl.setAutoFocusMode(dai.CameraControl.AutoFocusMode.OFF)
    ctrl.setAutoFocusTrigger()
    rgb_control_queue.send(ctrl)

    dai_now = dai.Clock.now()
    utc_now = datetime.datetime.now(datetime.timezone.utc)
    capture_start_time = utc_now
    dai_clock_start_time = utc_now - dai_now

    try:
        processing_loop(oak_device)
    except KeyboardInterrupt:
        logger.info("Exiting at user request")
        oak_device.close()
        logger.info("Camera Disconnected")
    except RuntimeError as e:
        utc_now = datetime.datetime.now(datetime.timezone.utc)
        exec_duration = utc_now - capture_start_time
        logger.error("Ran for %f seconds and then encountered a run time error",
                     exec_duration.total_seconds())
        raise e



if __name__ == "__main__":
    main()

jakaskerl

Hi @danl5
Could you minimize the code please; remove any parts that aren't necessary in reproducing the issue?

Thanks,
Jaka

danl5

jakaskerl

I experimented a bit more and I can reproduce the problem if the RGB and Depth streams are enabled (problem still occurs even with IMU turned off). but the problem seems to disappear if Depth is not enabled. Here is a pared down script that still causes the problem. (Changing the constant on line 15 from True to False allows things to run seemingly without issue).

import datetime
import depthai as dai

# Queue Names and Config
QUEUE_SIZE_MULTIPLIER = 4
COLOR_QUEUE = 'rgb_video'
COLOR_CONTROL_QUEUE = 'rgb_control'
DEPTH_QUEUE = 'depth'

# Color Config
COLOR_SIZE = (3840,2160)
COLOR_FPS = 12
COLOR_CODEC = dai.VideoEncoderProperties.Profile.H265_MAIN
COLOR_BITRATE = 20000000 # 20 Mbps
SET_WARP_MESH = True # whether or not to set the warp mesh for onboard undistort

# Stereo Depth Config
MONO_RESOLUTION = dai.MonoCameraProperties.SensorResolution.THE_720_P
MONO_SIZE = (1280,720)
MONO_FPS = 12

# Logging Config
DEPTHAI_LOG_LEVEL = dai.LogLevel.DEBUG

pipeline = dai.Pipeline()

#########################
# RGB Setup
#########################
cam_rgb = pipeline.create(dai.node.Camera)

# Properties
cam_rgb.setBoardSocket(dai.CameraBoardSocket.CAM_A) # CAM_A is RGB Camera
cam_rgb.setSize(COLOR_SIZE)
cam_rgb.setFps(COLOR_FPS)

# Define sources and outputs
control_in = pipeline.createXLinkIn()
control_in.setMaxDataSize(1)
control_in.setStreamName(COLOR_CONTROL_QUEUE)
control_in.out.link(cam_rgb.inputControl)

video_encoder = pipeline.create(dai.node.VideoEncoder)
video_encoder.setDefaultProfilePreset(COLOR_FPS, COLOR_CODEC)
video_encoder.setRateControlMode(dai.VideoEncoderProperties.RateControlMode.CBR)
video_encoder.setBitrate(COLOR_BITRATE)

rgb_out = pipeline.createXLinkOut()
rgb_out.setStreamName(COLOR_QUEUE)
cam_rgb.video.link(video_encoder.input)
video_encoder.bitstream.link(rgb_out.input)

if SET_WARP_MESH:
    print("Configuring RGB camera for on-board undistortion")
    cam_rgb.setMeshSource(dai.CameraProperties.WarpMeshSource.CALIBRATION)
else:
    print("RGB camera will NOT be configured for on-board undistortion")


rgb_out.input.setBlocking(False)
rgb_out.input.setQueueSize(COLOR_FPS * QUEUE_SIZE_MULTIPLIER)

#########################
# Stereo Depth Setup
#########################
left = pipeline.createMonoCamera()
right = pipeline.createMonoCamera()

left.setResolution(MONO_RESOLUTION)
left.setBoardSocket(dai.CameraBoardSocket.CAM_B) # CAM_B is Left Camera
left.setFps(MONO_FPS)
right.setResolution(MONO_RESOLUTION)
right.setBoardSocket(dai.CameraBoardSocket.CAM_C) # CAM_C is Right Camera
right.setFps(MONO_FPS)

stereo = pipeline.createStereoDepth()
left.out.link(stereo.left)
right.out.link(stereo.right)

stereo.setDefaultProfilePreset(dai.node.StereoDepth.PresetMode.HIGH_DENSITY)
stereo.initialConfig.setMedianFilter(dai.StereoDepthProperties.MedianFilter.KERNEL_3x3)
stereo.setExtendedDisparity(True)
stereo.setLeftRightCheck(True)
stereo.setDepthAlign(dai.CameraBoardSocket.CAM_A)

stereo.setOutputSize(*MONO_SIZE)

depth_out = pipeline.createXLinkOut()
depth_out.setStreamName(DEPTH_QUEUE)
stereo.depth.link(depth_out.input)
depth_out.input.setBlocking(False)
depth_out.input.setQueueSize(MONO_FPS * QUEUE_SIZE_MULTIPLIER)

#########################
# Start Oak Device
#########################
oak_device = dai.Device(pipeline)
oak_device.setLogLevel(DEPTHAI_LOG_LEVEL)
oak_device.setLogOutputLevel(DEPTHAI_LOG_LEVEL)

rgb_control_queue = oak_device.getInputQueue(COLOR_CONTROL_QUEUE)
rgb_video_queue = oak_device.getOutputQueue(COLOR_QUEUE)
depth_queue = oak_device.getOutputQueue(DEPTH_QUEUE)

ctrl = dai.CameraControl()
ctrl.setAutoFocusMode(dai.CameraControl.AutoFocusMode.AUTO)
rgb_control_queue.send(ctrl)

utc_now = datetime.datetime.now(datetime.timezone.utc)
capture_start_time = utc_now

try:
    while True:
        oak_device.getQueueEvent([COLOR_QUEUE, DEPTH_QUEUE],
                                    datetime.timedelta(seconds=2))

        rgb_packet = rgb_video_queue.tryGet()
        depth_packet = depth_queue.tryGet()

except KeyboardInterrupt:
    oak_device.close()
except RuntimeError as e:
    utc_now = datetime.datetime.now(datetime.timezone.utc)
    exec_duration = utc_now - capture_start_time
    print("Ran for %f seconds and then encountered a run time error" %
                    exec_duration.total_seconds())
    raise e

danl5

Furthermore, it seems that even without enabling extended disparity, left-right check and RGB depth alignment, the problem still occurs (lines 82-84). Same when I comment out the autofocus (lines 105-107).

danl5

Also, here is the debug output from running the pared down script:

Configuring RGB camera for on-board undistortion
[18443010E18C130900] [192.168.1.195] [9.339] [StereoDepth(6)] [info] Depth baseline: '7.5000005' cm, focal length: '844.2523'
[18443010E18C130900] [192.168.1.195] [9.339] [StereoDepth(6)] [info] Depth horizontal FOV: '74.32914', vertical FOV: '46.207264'
[18443010E18C130900] [192.168.1.195] [9.339] [StereoDepth(6)] [info] Depth intrinsics: 
 844.252,    0.000,  627.145, 
   0.000,  843.860,  367.051, 
   0.000,    0.000,    1.000, 
[18443010E18C130900] [192.168.1.195] [9.395] [VideoEncoder(2)] [debug] Bitstream frame size: 12604928, num frames in pool: 4
[18443010E18C130900] [192.168.1.195] [9.694] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 64.83 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 515 MB/s
[18443010E18C130900] [192.168.1.195] [9.694] [system] [info] Temperatures - Average: 43.50C, CSS: 44.87C, MSS 43.50C, UPA: 42.35C, DSS: 43.27C
[18443010E18C130900] [192.168.1.195] [9.694] [system] [info] Cpu Usage - LeonOS 43.88%, LeonRT: 38.75%
[18443010E18C130900] [192.168.1.195] [10.695] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2381 MB/s
[18443010E18C130900] [192.168.1.195] [10.695] [system] [info] Temperatures - Average: 45.04C, CSS: 46.00C, MSS 45.32C, UPA: 43.73C, DSS: 45.09C
[18443010E18C130900] [192.168.1.195] [10.695] [system] [info] Cpu Usage - LeonOS 64.56%, LeonRT: 7.78%
[18443010E18C130900] [192.168.1.195] [11.697] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2407 MB/s
[18443010E18C130900] [192.168.1.195] [11.697] [system] [info] Temperatures - Average: 45.43C, CSS: 47.13C, MSS 44.64C, UPA: 44.64C, DSS: 45.32C
[18443010E18C130900] [192.168.1.195] [11.697] [system] [info] Cpu Usage - LeonOS 62.03%, LeonRT: 7.59%
[18443010E18C130900] [192.168.1.195] [12.698] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2326 MB/s
[18443010E18C130900] [192.168.1.195] [12.698] [system] [info] Temperatures - Average: 45.77C, CSS: 47.80C, MSS 45.77C, UPA: 44.64C, DSS: 44.87C
[18443010E18C130900] [192.168.1.195] [12.698] [system] [info] Cpu Usage - LeonOS 58.78%, LeonRT: 7.57%
[18443010E18C130900] [192.168.1.195] [13.699] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2317 MB/s
[18443010E18C130900] [192.168.1.195] [13.699] [system] [info] Temperatures - Average: 45.89C, CSS: 47.13C, MSS 46.00C, UPA: 44.64C, DSS: 45.77C
[18443010E18C130900] [192.168.1.195] [13.699] [system] [info] Cpu Usage - LeonOS 60.14%, LeonRT: 7.54%
[18443010E18C130900] [192.168.1.195] [14.700] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2317 MB/s
[18443010E18C130900] [192.168.1.195] [14.700] [system] [info] Temperatures - Average: 45.89C, CSS: 47.58C, MSS 45.77C, UPA: 44.64C, DSS: 45.55C
[18443010E18C130900] [192.168.1.195] [14.700] [system] [info] Cpu Usage - LeonOS 59.91%, LeonRT: 7.43%
[18443010E18C130900] [192.168.1.195] [15.703] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2354 MB/s
[18443010E18C130900] [192.168.1.195] [15.703] [system] [info] Temperatures - Average: 46.00C, CSS: 47.58C, MSS 45.55C, UPA: 44.87C, DSS: 46.00C
[18443010E18C130900] [192.168.1.195] [15.703] [system] [info] Cpu Usage - LeonOS 61.61%, LeonRT: 7.68%
[18443010E18C130900] [192.168.1.195] [16.705] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2338 MB/s
[18443010E18C130900] [192.168.1.195] [16.705] [system] [info] Temperatures - Average: 46.17C, CSS: 48.47C, MSS 45.55C, UPA: 45.09C, DSS: 45.55C
[18443010E18C130900] [192.168.1.195] [16.705] [system] [info] Cpu Usage - LeonOS 61.27%, LeonRT: 7.56%
[18443010E18C130900] [192.168.1.195] [17.707] [system] [info] Memory Usage - DDR: 250.81 / 333.39 MiB, CMX: 2.47 / 2.50 MiB, LeonOS Heap: 65.08 / 82.31 MiB, LeonRT Heap: 9.07 / 40.50 MiB / NOC ddr: 2342 MB/s
[18443010E18C130900] [192.168.1.195] [17.707] [system] [info] Temperatures - Average: 46.00C, CSS: 47.58C, MSS 45.55C, UPA: 44.41C, DSS: 46.45C
[18443010E18C130900] [192.168.1.195] [17.707] [system] [info] Cpu Usage - LeonOS 60.95%, LeonRT: 7.59%
[18443010E18C130900] [192.168.1.195] [1710260093.952] [host] [warning] Monitor thread (device: 18443010E18C130900 [192.168.1.195]) - ping was missed, closing the device connection
[18443010E18C130900] [192.168.1.195] [1710260095.952] [host] [debug] Log thread exception caught: Couldn't read data from stream: '__log' (X_LINK_ERROR)
[18443010E18C130900] [192.168.1.195] [1710260095.952] [host] [debug] Timesync thread exception caught: Couldn't read data from stream: '__timesync' (X_LINK_ERROR)
Ran for 21.547966 seconds and then encountered a run time error
Traceback (most recent call last):
  File "./onboard_undistort_no_imu.py", line 127, in <module>
    raise e
  File "./onboard_undistort_no_imu.py", line 117, in <module>
    rgb_packet = rgb_video_queue.tryGet()
RuntimeError: Communication exception - possible device error/misconfiguration. Original message 'Couldn't read data from stream: 'rgb_video' (X_LINK_ERROR)'
[18443010E18C130900] [192.168.1.195] [1710260095.960] [host] [debug] Device about to be closed...
[18443010E18C130900] [192.168.1.195] [1710260095.960] [host] [debug] RPC error: Couldn't write data to stream: '__rpc_main' (X_LINK_ERROR)
[18443010E18C130900] [192.168.1.195] [1710260095.960] [host] [debug] shutdown call error: Device already closed or disconnected: Input/output error
[18443010E18C130900] [192.168.1.195] [1710260097.952] [host] [debug] Getting crash dump...
[18443010E18C130900] [192.168.1.195] [1710260115.938] [host] [warning] Device crashed, but no crash dump could be extracted.
[18443010E18C130900] [192.168.1.195] [1710260119.933] [host] [debug] Device closed, 23973

jakaskerl

HI @danl5
Just letting you know I'm looking at the issue.

danl5

Any luck?

jakaskerl

Hi @danl5
I have forwarded the issue to dev team. Looks like a FW issue regarding Camera node stability, but not sure still. I will reply here when we have some conclusive evidence/solution.

Thanks,
Jaka

danl5

@jakaskerl any change here?

jakaskerl

danl5
There have been several updates with regards to both, the overall stability as well as the Camera node.

Are you still experiencing the same issues on newest bootloader (0.0.28) and depthai 2.28?

Thanks,
Jaka