Realsense ZR300 rosnode: IR & RGB are not timestamped properly about maplab HOT 7 CLOSED

@linhuincu So the error message you are seeing means that the ROS node that talks to the sensor is unable to get infrared and RGB images that have been timestamped with the microcontroller clock. This call should have told your ZR300 that you would like to have all data properly timestamped, but it looks like that's not happening on your device. Do you happen to have an early version of the ZR300?

I'm not sure how to solve this, if it can be solved at all without tinkering with the sensor software. However, if you are only interested in running state estimation on the ZR300 you can simply turn of IR, RGB and depth publishing in the launch file (here and below). Alternatively you can make a small change to the ROS node by commenting out this check such that it publishes the IR/RGB/Depth data even if they are not properly timestamped.

from maplab.

@mfehr - I'm having a similar issue with my ZR300 as well, though with the fisheye timestamp. I imagine this will result in poor VIO results using ROVIOLI.

When launching maplab_realsense.launch I get the following error:
E0708 16:50:47.415486 19441 zr300.cc:568] The timestamp of frame 215 of type FISHEYE did not originate from the motion tracking microcontroller. The timestamp domain is set to: 0 (camera = 0, microcontroller = 1). Skipping frame!

Any suggestions on troubleshooting this issue?

Thanks!

from maplab.

@bsbretly does this happen for all frame types and all frames or only once in a while?

from maplab.

@dymczykm this only happens once in a while.

from maplab.

@bsbretly that shouldn't matter that much then. The wrongly timestamped frames are rejected, but all the others are OK and consumed by the estimator. I don't think this would cause much difference in the VIO output.

from maplab.

Thanks for the help @dymczykm. I tested this earlier today and it seems to estimate pose quite well despite the error.

Appreciate your input!

from maplab.

hi @mfehr
yesterday i have the similar error and i already check your suggest link to see the code but not really know what your mean i think maybe the line you marked is already change from the new update.
i was opened a new issue of my error proplem at here
https://github.com/ethz-asl/maplab/issues/152

from maplab.