Open osrf-migration opened 6 years ago
Original comment by Stefan Kohlbrecher (Bitbucket: Stefan_Kohlbrecher).
Reading PR #2502 it looks like this is expected, but not intended behavior :) Can that PR be merged?
Original comment by Nate Koenig (Bitbucket: Nathan Koenig).
I think we can get pr #2502 merged. We are having a PR review sprint the first week of December, and we'll aim to get this wrapped up before or during that week.
Original report (archived issue) by Stefan Kohlbrecher (Bitbucket: Stefan_Kohlbrecher).
This is on Ubuntu 16.04/Kinetic/Gazebo9 from .debs.
I have a fairly complex world (which I unfortunately cannot share) and a (ROS-based) robot with 4 cameras (using
libgazebo_ros_camera
) moving inside it.As long as I only subscribe to the image of one of the cameras, everything largely works fine. I get a realtime factor of ~0.95, images come in at a rate of 10Hz (as specified). When subscribing to the images of the three other cameras, the following effects are observed:
ros::Time::now()
inside a ROS image callback, the delay is about 1 (simulated) second.It thus looks like rendering images takes considerable time. That's ok in principle, but what I'd expect in that case is that Gazebo slows down accordingly to keep the image publishing rate. Is the observed behavior the intended/expected one?