Does anyone know how we would be able to use two cameras without exceeding the 7Mb/s limit? Last year we used two cameras on our robot but ended up removing one of the cameras because we were managing to use way too much bandwidth at the time when there was no limit. Now we are stumped but we would love to have one camera on our disc intake and another on our shooter. I’ve considered having a switch that changes which camera is initialized at a time but worry that this will cause the camera to have to reboot every time we switch?
Turn the compression setting way up (60+).
It also helps to turn the resolution down. Ignoring compressions, a 640x480 image has about 310 thousand pixels, a 320x240 has 1/4th as many, and a 160x120 has 1/16th as many. Compression impacts it as well. The default dashboard will tell you an approximate bandwidth used by the camera. If you add a second camera using a LV dashboard, just add the two numbers together before displaying.
Greg McKaskle
I think your problem last year may have been exceeding the processing power of the classmate with two cameras, rather then overruning the field bandwidth. This year, the charts tab on the driver station shows the Driver Station CPU %, which will make it easier to track down those kind of issues.
Thanks for the suggestions, we will be sure to try them out. We shouldn’t have had a problem with CPU because we opted for a more powerful netbook last year in place of the Classmate.
Sorry, I might have been thinking about a previous year.
Just reduce the settings. To do it, log onto your camera’s web interface. Go into settings and change the resolution/quality/compression. For you, you might want to reduce the frame rate, and reduce the resolution a bit. put the compression level at 40-50. That will give a good enough quality to be usable.
- Increase the Compression
- Decrease resolution
- Decrease Frames Per Second
Hope that helps!
Frames per second is probably the first thing you should try reducing, with compression set to 30 to start with. Find the lowest fps value that is still useful to your drivers. Computer vision isn’t likely to require more than 4-8 fps. Then reduce the resolution to the lowest value that still works for you, and then if you’re still close to the limit increase compression.