Base stations 2.0 disappearing An trackers losing tracking

[going2kilzu]

Hi 

So I recently bought a set of base stations 2.0 an 3 trackers an they have bin working fine for the past 2 days.  Today I got a set of mounts to put them on my wall.

They worked for awhile but then I noticed it was losing tracking.

So I tried to figure out the issue but I cant seem to find how to fix it.

Ill put in a video showing what its doing at the moment.

[HackPerception]

@going2kilzu, Can you upload a SteamVR log via, SteamVR -> Generate System Report. Save it as a .txt.

My guess based off of the video alone is that you have a reflection issue or that specific  basestation that you're pointing towards has an internal fault or is seeing power fluctuations. Try replicating the issue without the headset on and see if that basestation's LED stats lights change.

You might not have the stations mounted securely enough and the vibration is too great. The speed at which the station picks back up supports that theory.

[going2kilzu]

On 1/11/2021 at 2:16 PM, VibrantNebula said:

@going2kilzu, Can you upload a SteamVR log via, SteamVR -> Generate System Report. Save it as a .txt.

My guess based off of the video alone is that you have a reflection issue or that specific  basestation that you're pointing towards has an internal fault or is seeing power fluctuations. Try replicating the issue without the headset on and see if that basestation's LED stats lights change.

You might not have the stations mounted securely enough and the vibration is too great. The speed at which the station picks back up supports that theory.

Sorry for the late reply I had gotten a little busy

At the moment its only with the one station above in the video.  I have another set an it works just fine.  It only seems that when I turn to look at the one above it freaks out.  The hip an only the hip will fly away an completely freak out when in use but the feet stay solid.  But once I unplug the one an only use the other station I had no issues running it for a full play session.

regarding the mount its as solid as it will get.  I had t o buy mounts separate due to the standalone 2.0's not coming with mounts since I run a hybrid set up.

its very weird cuz sometimes it happens an sometimes it doesent.  

 

The weird part is it only seems to have a problem when it is located on my waist Here in this video I bounce from one station to the next.  the first standing ide position is the station im having trouble with the the one when the front of the Puck is facing the came is the one I have no problems with when in standalone 

SteamVR-2021-01-14-PM_11_05_04.txt

Edited January 15, 2021 by going2kilzu

[HackPerception]

@going2kilzu I suspect you have an environmental reflection problem. In the logs, there are steady reports of "back-facing hits". In 25 seconds of sample logs below, I see 4 different callouts about reflections. It's not counting a particularly high number of back-facing hits but definitely enough to cause a device's pose estimate to wonk out hardcore. The low but steady count leads me to guess that you have 1 specific trouble surface in your room. Two of the callouts also list a decently high velocity - might be correlated; could be that you're not getting enough good optical samples and the high velocity is greater than what the IMU can make up for.

IR light is a real PITA to try and troubleshoot reflections with because humans can't see in IR and materials reflect IR differently than optical light so it's non-intuitive. Using a flashing cellphone light or headlamp remains one of the more reliable ways to try and track it down. Any shiny surface can be a problem.

 

[going2kilzu]

I dont really have much in regards a reflective surface other then like 1 photo on a shelf.  Since I run an oculus with the stations I could probably try using the IR cams in the headset with passthrough to see what its hitting.  But other then that I wouldnt know since one of them works just fine but in unisons they freak out

9 hours ago, VibrantNebula said:

@going2kilzu I suspect you have an environmental reflection problem. In the logs, there are steady reports of "back-facing hits". In 25 seconds of sample logs below, I see 4 different callouts about reflections. It's not counting a particularly high number of back-facing hits but definitely enough to cause a device's pose estimate to wonk out hardcore. The low but steady count leads me to guess that you have 1 specific trouble surface in your room. Two of the callouts also list a decently high velocity - might be correlated; could be that you're not getting enough good optical samples and the high velocity is greater than what the IMU can make up for.

IR light is a real PITA to try and troubleshoot reflections with because humans can't see in IR and materials reflect IR differently than optical light so it's non-intuitive. Using a flashing cellphone light or headlamp remains one of the more reliable ways to try and track it down. Any shiny surface can be a problem.

 

 

[going2kilzu]

ok so update I tested by removing anything remotely reflective other then my monitors an I get the same result with the one station.

At this point im about 80% sure its just the station its self

[going2kilzu]

I think I might have fixed it.  I tried moving around my monitors to see if maybe it could be bouncing off those.  ATM its working but ill give it alittle to see if it continues 

[HackPerception]

@going2kilzu - I'd also triple check to ensure that the base stations are on different stations. The device could have theoretically reset itself to channel S-1 and if Bluetooth isn't working you might not see an error message . If you have a bad station, the firmware in it should notice the error and stop it from spinning up.

I'd also double check that the firmware is recent on all devices. If you have older firmware on one of the devices - it could cause problems.

There was deffo a reflective surface in the environment per the logs. It wasn't causing the number of errors you'd see with something like a mirror, but definitely enough cause momentary tracking loss.