Who has stable LH 2.0 tracking on 8Kx 😀

[Darrell]

Please can someone tell me where i can get the jitter software, or send a link?
I am in China and can not access the google drive link.

[Chumet]

You have to use only one base station on channel “B”, turn off the other. y Turn and move the hmd away, more than 2 meters.

[Chumet]

I was having trouble getting base station 1.0 to detect all sensors. I put it 1 meter high and about 2.5 meters away. Then use: lighthouse_console.exe -> sensorcheck, and see moving the HMD in front of the lighthouse to check that all the sensors are showing hits (HitCount). Otherwise vrtrackingcalib.exe will fail.
Of course the rest of base stations 2.0 turned off as well as any other tracking device, knuckles, etc. Ahh and the Pimax off only with the red Led on. Without Piservices or Pitool.
Before starting with this whole issue of calibration, I had bought 2 base stations 1.0 thinking that they were the solution for a stable monitoring and I realized that they are super sensitive to the environment and the HMD, I could never get it to position itself. After researching the calibration, I used only one of them as recommended to do the calibration and I have returned them. The 2.0 are much more compatible with our Pimax and with reduced and complicated gaming environments, as is my case.

Try and do not be discouraged,

[Trigen]

Guess its a good idea to move the HMD around so it will actually calibrate.

[Trigen]

Sadly the calibration didnt do much. The rotational tracking is still terrible and it when i look straight left i will get a shift in position which is really annoying. I have 2 1.0 LH in each corner of the room about 3 meters away looking 45 degree down

Edit. i see Pos dev is 145 and Rot Dev 1450 after a room setup again but thats still terrible

[Chumet]

Hello @Trigen, at the end of the calibration I assume that you have sent the .json that has been generated to the HMD, right?

Using:

lighthouse_console uploadconfig auto_LHR-3499941F.json
And you have deleted “previous backup” .json from these routes:
C: \ ProgramData \ pimax \ runtime \ backup \ lighthouse (LHR-3499941F_config.json)
C: \ ProgramData \ pimax \ runtime \ config \ lighthouse \ lhr-3499941F (config.json)
?: \ Steam \ config \ lighthouse \ lhr-3499941F

It is also important to delete:
C: \ ProgramData \ pimax \ runtime (chaperone_info.vrchap)
?: \ Steam \ config (chaperone_info.vrchap)

And do the room calibration again in Pitool and SteamVR, but taking into account:

Place the HMD in a position that is highly visible to both base stations.
Do not touch it or interfere between the base stations and the HMD and confirm the direction and then the height (if you have not put it on the ground, write the height yourself).
And without touching it from SteamVR it calibrates as sitting or standing and reconfirms the direction and height.

I think it is important not to touch the HMD of its position during the two calibrations.

I think there are problems between the calibration that Pitool does and SteamVR. Collecting different data can cause jitter.

You can also test, after all this, if with a single base station the jitter is better.

In my case, before calibrating my Pimax, I had that follow-up jump when I turned on myself, there were moments that everything began to shake, it deviated and then when I continued turning it returned to its position. It is all the fault of the fact that there are specific positions of the HMD in the room that the sensors do not determine correctly or with positioning errors. It may be because at that point there are few sensors in sight of the base stations, or they are not well calibrated.

If you want to send me your original config.json and the one that you managed to generate after calibration, I can easily compare them and see if there is much difference.

[walz]

Hey guys,
This seems like a good place to ask.
I’ve been having tracking problems recently. One of my V1 base station died, which I noticed by the fact that I had a dead spot where I’d loose tracking for all devices (controllers or HMD). It went worse and worse.
Towards the end I started to experience one specific issue in blade and sorcery. When i turned my head towards the dead spot, the character would turn 180 but then if I continued turning it would resumed normal positionning (the dead spot was positionned in my 4 o’clock or so). Kind of hard to explain, but it was really jarring and made the game unplayable.
I found some second hand V2 base stations for sale and switched them out. The tracking is mostly solid, I get some occasionnal jitter but it could live with it.
However the issue in B&S persists… I tried redoing Room setup both in pimax and steam vr to no avail. Reinstalled the game, deleted the config files, the lighthouse config files in program data/pimax/runtime but nothing works.
I ran the tools I found in this thread and got this :

Jitter test result is pretty average, It ranges from 1.0 to 1.5.

Do you guys thinks I should try the recalibrating you guys have been talking about ?

[Chumet]

If you’ve been fine with lighthouse v1.0, it’s probably not a question of calibrating the optical sensors. By the way, what HMD do you have? 8KX or other with blue rubber coating? It seems the tendency to cause jitter, possibly due to the dispersion of factory calibrations, at least it was my case. Still having a lighthouse v1.0 in your possession you can do a calibration, but more as an experiment and by knowing the status of your factory calibration, it may help you to improve, but it is to test it. There are others that the calibration has not helped them, since the problem would be another.
What I can assure you is that the lighthouse v2.0 are very robust in not losing track because of the way they send the synchronization data to the HMD, that makes them always positioned although there is dispersion in the real position. causing jitter. On the other hand, the V1.0 or are perfectly synchronized with each other and with the HMD or the follow-up is lost. That happened to me when I wanted to switch from V2.0 to V1.0. There was no way for me to position the HMD, when I got it, any movement would lose track again. Due to the miscalibration of my HMD. So as soon as I had it properly calibrated and working perfectly with both V1 and V2. I forgot the V1.0.
Before getting into calibrations, try looking for options for positioning the lighthouses in your room, distances, heights, inclinations of 30 to 45º.
You can also check if all the sensors of your HMD are visible to the lighthouse permanently or there are some that have a hard time getting hits. This would be with: lighthouse_console.exe -> sensorcheck. And moving the HMD in all its angles where there are sensors, there are 32 in total and the deviation of any of them will cause jitter when the position of the HMD sees that sensor and few more. That was my case.

[Chumet]

Taking a good look at your screenshot of the trackingstats, I see that the positioning of the controllers is just as bad as that of the HMD, it should have much better figures in the controllers if it were a calibration problem. Try trackingstats by leaving both the HMD and the controllers in an area that is highly visible to the lighthouses, and don’t interfere with their presence. Leave them 10000 frames and compare results.

[walz]

Tanks for taking a look. I appreciate it.
I have a 8kX so indeed with the blue coating.
I’ve run the tracking stats again with both controllers and hmd on the ground in the center of my play space and this is what I got :

I’ve run the sensorcheck and couldn’t identify a problem.
What is weird is that the tracking is stable, I do occasionnally get an “earthquake effect” for a second or 2 except its only in B&S where I get that 180° jump and back. Also, it happens independently of my position in the room, whether I’m in the center, or in a corner, crouching or whatever, tracking is fine but when I turn my head to the 4o’clock position, the camera will do a 180 and then resume tracking normally.

Edit: So I took out my remaining good V1 lighthouse and put it up and I still get the same problem. So definitely not a V2 problem, might be a problem with the game…
Anyway thanks for your help Chumet

[Chumet]

Actually the data you show is good now, and has the same deviation as the controllers. You shouldn’t have that follow-up jump when you’re at 4 o’clock though. I think you may have a badly calibrated sensor, but it doesn’t seem important. If you want, send me the config.json file of your Pimax and he could buy it to see if there is any sensor out of the correct position.
It is in the path: C: \ ProgramData \ pimax \ runtime \ config \ lighthouse \ lhr-xxxxxx, you will have several folders of that style that belong to the controllers and the pimax, it is easy to know what it is, opening the config.json you will see almost at the bottom “manufacturer”: “Pimax”, and in the others it would be something like this: “manufacturer”: “Valve”,
“model_number”: “Knuckles Right”

You tell me something and maybe we can fine tune the config.json to have the perfect tracking.

[walz]

Ok I’ll upload the json to google drive and send you the link via MP.

[Chumet]

Hello @Walz, you have several sensors with a lot of dispersion more than 2mm, that comparing it with my calibrated point cloud like 10 times, but it works for me with a very very solid tracking, so I deduce that any other Pimax that works correctly does not It should have so much deviation between sensors, the funny thing is that most if they coincide with mine. You can start easy and it is to use my already modified config.json so that it works for you in your Pimax. I leave it here public in case it helps someone. But the ideal is to calibrate it yourself several times and compare results until you see that more or less you always have the same position in all the sensors. As that is already more expensive, start by using the config.json that I put here. I discovered this when I asked another fellow on the forum for the config.json who claimed to have a good following and it was a spectacular change.

In green: Chumet
In blue: Walz

https://drive.google.com/file/d/1_LDGkVQ-KpMHEe4xNgmfFoUtXCqzNyDv/view?usp=sharing

To send it to Pimax you have to put the file in this path and then execute from CMD:

?:\Steam\steamapps\common\SteamVR\tools\lighthouse\bin\win64
lighthouse_console uploadconfig config.json

Make a backup of your config.json and delete it from the folder with which you got it, calm that there will appear again but the one that you have previously sent to Pimax, as soon as you start Pitool.
C: \ ProgramData \ pimax \ runtime \ config \ lighthouse \ lhr-xxxxxx

[risa2000]

@Chumet

To send it to Pimax you have to put the file in this path and then execute from CMD:

?:\Steam\steamapps\common\SteamVR\tools\lighthouse\bin\win64
lighthouse_console uploadconfig config.json

I would strongly suggest to anyone, who wants to change their optical sensor positional data in the JSON config to:

• First backup your actual config, by using lighthouse_console.exe this way:

lighthouse_console.exe downloadconfig

while the headset is attached. This will create a file LHR-<your_serial_number>.json in the same directory.

• Save this file.
• Replace the section called lighthouse_config with the new config, you want to test. e.g. the one from @Chumet.
• Upload the new config to the headset by using:

lighthouse_console.exe uploadconfig <name of the file>

• (EDIT 15.4) You need to power off/on the headset (to reboot the lighthouse controller inside the headset), in order to have these new values taken on.

This way, you will preserve the serial number in the config and, which is far more important, the IMU calibration data (IMU bias and IMU scale).

In general however, I am sceptical about this approach, because you are basically trying some random calibration on some random headset. It may change things (either way), but only by pure luck.

[flynny75]

@PimaxUSA if only there was a way we could calibrate ourselves with LH 2.0.

Please try and get the newer calibration tool.

[thenic123]

Can Steam Valve 2.0 Base Station work with HTC 2.0 Base Stations?

Would have better tracking and less jitter if I have two 2.0 base stations? Where would I mount the 2nd BS?

[Heliosurge]

There the same Lighthouses. There are from my understanding the v2 SteamVR (Which is first release) and the revised version that says Valve Lighthouse v2.0. Only Valve makes v2 lighthouses. No one else is licensed to make them.

[risa2000]

Yep, different hardware revisions, but they are compatible, i.e. both can be used at the same time.

[walz]

Hey @Chumet @risa2000, thank you both.
I uploaded your config file using risa2000 advice, redid the pimax and steamvr room setup and it’s looking good so far.
The jittertest improved and I didn’t experience any “earthquake” effect so far. Might be just luck or placebo effect but I’ll leave it at that for a while to see if it’s stable.

[levers]

I have to say, now I’ve finally received my Kickstarter lighthouse, that it was not worth the wait. I put it on the wall mount where I have my V1 and it constantly jumps my position then drifts back to where it should be. I tried moving it which stopped the jumping but using the motion compensation for sim racing the car moves to the side when I go up or down hill. Not sure why after all this time I have not got the later version of the V2. Would that make a difference?

There’s always eBay!