Pimax 8KX issues with lighthouse tracking

Pimax Hardware P2: 8kX
1

Hey Everyone - i’m having an issue with lighthouse tracking lately. About 6 months ago, I finally was too frustrated by the 1.0 tracking issues so I upgraded to 4 2.0s. I have them setup in the two corners of my room in front of the direction I face and one in the rear right corner. I also keep one on the window sill in front of me pointing directly at my headset with my thought being, I can’t possibly have an issue with a lighthouse only 3 feet and right in front of my face.

Initially, this setup seemed to work amazingly, but as it always seems with odd things in Pimax world, issues started to crop up. Initially, I had near instantaneous tracking pick up and now it sometimes takes a few minutes to show up. Additionally, while racing in ACC, when I look down and to the left to see the relative/pitscreen/whatever, I will almost always get a screen jump. Sometimes it happens looking in other directions, but most notably looking downwards and left. This is despite having clear line of sight to the entire headset from the top left, front middle, and top right lighthouses.

In the last few days, I’ve done quite a bit of troubleshooting to try and figure this out on my own, and with the help of the lighthouse_console.exe in the pimax runtime folder and some of the sub commands, most notably sensorcheck, I was able to figure out a few things:

  1. A lighthouse 3 feet in front of your face actually sort of sucks compared to the ones that are about 10 feet away and up on the wall.

  2. According to the sensorchecks data, sitting still in my race seat with my hands on the wheel, I get roughly 70 to 90% of the maximum connectivity rating from 3 out of 4 of my lighthouses with only the rear right being fair low. This makes sense at it would barely catch the right right corner unless I turned my head 90 degrees towards it.

  3. Any time I raise either my left hand or right hand to grab the corresponding side of my headset to adjust or lift, about 90% of the time, tracking will start drifting massively with the left hand and maybe get skippy on the right. When comparing this against the sensorcheck command, I lose roughly 90% of the signal with my right hand touching the right corner and roughly 99% with my left hand touching the left side. This is despite almost the entire headset being clearly visible to the lighthouses.

I believe lighthouses communicate over Bluetooth to the headset and as best I can tell, I’m somehow blocking that connection by touching the headset. I suppose I could also be “grounding” an antenna if it is inside the headset and mounted wrong. I was hoping someone else had experience with this or could offer any further suggestions with how to troubleshoot it.

2

I have four 2.0 basestations that I’m using with my Pimax 8KX, and they are working very well. I have them placed in each corner of my room, up top near the ceiling. This is the ideal placement for them.

As I think you’ve already figured out, placing one of them at point blank in front of you isn’t a good location. The primary issue with being so close is that the tracking area is so small at point blank. Basestations are able to work just fine from considerable distance with no loss in capability, and the further you get from them, the more area their FOV covers.

They do not. The basestations communicate via laser.

Basestation tracking is often incorrectly called “outside in” tracking. The basestations are not sensors. Rather they are laser emitters. The sensors are actually in the headset and controllers. They detect the laser beams coming from the basestations.

For your tracking problems, what you should look for are reflection issues. Mirrors, glass, computer monitors, shiny things, etc. In addition to obvious sources, some things that aren’t reflective in the visible spectrum are reflective in the IR spectrum.

You probably have a reflection problem, but the presence of multiple basestations is masking it. It’s able to still get the right answer much of the time, but it will be getting conflicting data.

Turn on only one basestation at a time and test its tracking. You’re likely to find that one of them is getting wonky results, and it’s been compensating with the other three. But sometimes that one becomes the only one it can see or it otherwise decides to accept its bad tracking data over the good data from the others.

This could be because one of the basestaions is bad, but more likely its because of reflection problems.

I’ve had such issues myself in the past. I’ve covered up various reflective surfaces in my playspace and that resolved the issues completely. Just start putting painters tape on everything shiny or covering it with cloth until your tracking problems go away.

1 Like
3

Hey Sargon - thanks for the response. I did actually do some testing with covering my reflective monitors and a TV I have and so far as the data quality according to the sensorchecks app from Pimax there was no difference. My biggest concern was in my normal playing position, my rear right was barely tracking unless I purposefully moved my headset towards/twist it. My thought was that as that thing drops enough it loses tracking and somehow the switch between enough tracking data and not enough from it is part of what is causing the glitchyness.

However, none of that explains why when touching one side of my headset or the other, and 95+% of the headset in clear view that the screens will lose tracking entirely or start drifting massively.

1 Like
4

Surprised you also had trouble with v1 bases. Mine have been solid for years. First i though reflections because you mentioned setting a base on a window sill but then it starts to sound like a problem in the 8KX. Maybe open a ticket.

1 Like
5

The V1s I had, installed perfectly, with all reflective surfaces covered and windows/blinds closed were awful. They worked OK for room side stuff, I think b/c the rear one could actually help, but for sitting in a racing seat, it was terrible.

1 Like
6

I don’t think I’d trust the app to indicate whether you will experience tracking errors. Especially if you’re checking for reflection problems. The whole thing with reflections is that it can’t tell which is direct and which is the reflection, so how will an app be able to tell you?

The app could be useful for additional clues about what’s going on, but I think primary testing should consist of moving your body and performing physical tests to see if tracking is actually working accurately or not.

And again I’ll emphasize that testing the basestations individually is important to narrowing down problems. The more basestations you have, the more the other basestations will mask problems that are only being experienced by a single basestation. If you isolate to just that one basestation, what’s going on with tracking failure can become obvious.

This could be an RF issue. Do you have a direct drive steering wheel? Fanatec direct drive in particular is known to cause heavy RF interference that can actually knock out tracking on Pimax and some other VR headsets which are plugged into the same PC.

You could test for this by completely powering off any steering wheels and actually I’d power off all nonessential electronics in the room and maybe even adjacent rooms to see if it has an effect on this. If the problem goes away, then you can start powering things back on to isolate which device is causing interference problems.

You could also look into your shoes and carpet. Maybe try an ESD strap and see if that has any effect.

Shortly after moving into my current house, I discovered that there was improper wiring and the entire ground was electrified. Weird stuff can happen.

1 Like
7

That is weird. If the bases were in the same place then seated performance shouldn’t be anymore an issue than moving around in roomscale. Heck, I found one V2 base sufficient for racing (i have both V1 and V2). Even found one V1 is solid for seated.

8

The front face of the 8k/5k has only something like four sensors - two per “facet”, in opposing corners, and most of the remaining sensors are on the sides, so covering one side with your hand occludes much more than you’d think…

9

Hey, I hear what you are saying, but lets say for instance I cover the left side (really just the first inch of the left side). The right side lighthouses should easily see the right side. Certainly the right rear can’t even see the left side, but tracking dies entirely. The sensorcheck app shows going from values of 1000+ (which are near peak) down to 0 to 20. I had one of my friends test his v2 (i can’t remember the actual number) 8kx and instead of losing 99% of his sensor strength, he lost about 10 to 20% when repeating the same covering test. There’s gotta be something else.

1 Like
10

I think occulsion due to the racing seat headrest and moving around inside the car has an effect. The angle from that corner to my head would at most show a small portion of the right rear side of the HMD to the lighthouse and I think that’s where it had issues. I’ve also hear that the v1s rely on infrared and even though I did my best with the blinds being closed, they couldn’t comletely stop all sunlight.

11

So what is still using RF between the headset? I said bluetooth, but tthat was just a guess at the type of connectivity. What is the headset using for RF between it and the lighthouses? I do use a fanatec DD but it is the new tiny one and I have tested with it off and on and still see similar results.

As far as not trusting the sensorchecks, its pretty much a 1 to 1, when the sensor strength dropps below about 200 on the available lighthouses, tracking starts jumping or drifting. If it is above that, tracking basically has no issues whatsoever. At peak values, it can get as high as 1100 to 1200.

I’ll try the one base station at a time. The sensorcheck app does check each for strength individually and the rear right one barely hovers above that 200 mark that seems about the limit of proper tracking and i think that’s simply the fact that it can only see a tiny portion of the headset. I’m wondering if it suddenly dropping below that 200 point is what is causing the jumping/drifting. It would make sense because I would think looking down and left would present an even smaller amount of the HMD to it. At the same time, it doesn’t make sense that the 3 other trackers which can all plainly see it just fine don’t make up for it. It would seem more of a software design error in caring about which lighthouse information is the most important.

12

RF in this case means that it’s actually knocking out the electronics. It’s not related to any sort of wireless transmission (the only transmission involved in this case is by laser, which is immune to RF). But I think what happens in the Fanatec DD wheel case is the sensors themselves which detect the laser beams are suffering so much electrical interference that they stop working.

This is a known issue for Fanatec DD, and you can find a lot of threads and stuff about it. Those things apparently put out a fantastic amount of RF which can affect sensitive electronics. Solutions seem to involve winding its cables through large ferrite magnets and moving its cables farther away from sensitive devices.

You mention that turning your DD wheel off doesn’t make a difference, so that would seem to kill that theory. But there might still be something else going on with some other device or something which is producing a similar effect. Or maybe a ground loop or something.

Of course, another possibility is that your 8KX is defective. What you’ve described is the kind of thing that could happen, for instance, if it has grounding problems internally.

Hmm. Or it could be related to the grounds on the ports its plugged into, or something like that.

Oh, here would be a good test. Is there another PC you can plug your 8KX into which is in a different environment? Especially if you could try it out at a friend’s house so it’s a different PC and different environment entirely. You could see if the problem follows the VR headset or if it’s something about the environment its in.

13

Yea - I have an 5k Super up in the attic I need to go get down and a/b test with. And I just remembered my kids computer used to be mine and has all the PiTool stuff just waiting for me to goof off with. I’ll test that out. Thanks for the idea!

EDIT: I thought it might have to do with some type of grounding issue as well. At least with the cover up and complete drop out. My thought was that whatever is on the left/right sid eis somehow connected and I’m groudning it out by touching the headset or having my hands very close to it.

14

I am sure you are right, but sensor placement is none the less one of the general factors in tracking coverage and quality (…and the p8k/5k has them where it does); One want at least three or four sensors which have both sight line to a given base station, and as much distance to one another as possible – the larger physical scale helps with resolution. :7

15

i have my V1 bases at diagonal to the center of my play space. (about 10 and 4 o’clock). both wall mounted at 8 foot elevation. Each 6 feet from center point down at that center, Although my chair isn’t high-back, a racing chair would not block line of sight either. Not at 8 feet high.
i have never experienced occlusion in racing and flight sims. Even when hastily checking my six for that persistent FW190 on my butt. Dam wingman. :grinning:

2 Likes
16

To reply for anyone looking in the future, I removed the lighthouse directly in front of me and the one to the rear right upper corner when I’m in my racing seat. IN the last 3 play sessions, I haven’t had a single case of screen jumping and the only issue I really have is that when i lift my headset up, i have to make sure I do it from the middle and not either side to avoid the screen drift and completely tracking loss. I have a ticket open with Pimax as my buddy’s 8KX (v2) doesn’t do this when lifting from the left or right side of the headset.

So the thing summarize is that the headset I have doesn’t deal well with trackers losing connectivity/signal (whatever you wanna call it) with the headset individually even though other lighthouses work just fine. I removed the two which had borderline bad signal and left 2 that had great signal from the start and continue to and suddenly my screen jumping is gone. So what i’ll do moving forward is only turn those other lighthouses on when I’m doing room scale VR. I’m not sure why the Pimax does so poorly with ignoring tracking data from lighthouses which are losing connectivity and instead using the data from the good ones, but I hope they patch the thing in a future firmware update.

17

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.