Who has stable LH 2.0 tracking on 8Kx 😀

cruser_icu · March 3, 2021, 10:18pm

Thick shag carpet

MrAhlefeld · March 3, 2021, 10:52pm

Using part of the SteamVR Tracking HDK (imu_calibrator.exe)
I found out that my IMU was incorrectly configured in the 8KX configuration file…

original values:
“imu” : {
“acc_bias” : [ 0.05108999830, 0.07593999799999999, 0.07740999760 ],
“acc_scale” : [ 0.9983000160, 0.9990000130, 0.9911000129999999 ],
“gyro_bias” : [ -0.008832000200000001, 0.02075999980, 0.001582000060 ],

Corrected values:
“imu” : {
“acc_bias” : [ 0.07382, 0.05502, 0.03354 ],
“acc_scale” : [ 0.07382, 0.05502, 0.03354 ],
“gyro_bias” : [ -0.007885, 0.0237, -0.007775 ],

Moving along… now if I just knew someone in Denmark with either an Index HMD or a Vive Pro, then I would be able to correctly update my basestations to newest Firmware.
Then I would be able to use: vrtrackingcalib.exe /usedisambiguation synconbeam
To capture the sensor points on my 8KX, this is why @Chumet needed to use a LH1.0 to do the same.
I’ve used part of his configuration file on my 8KX for the sensor points, and it has made a positive difference.
We are not quite there yet… I just wish Pimax would make progress faster than us

MrAhlefeld · March 3, 2021, 10:55pm

Positively NO reflections from your basestations in that

risa2000 · March 3, 2021, 11:45pm

The acc_bias might be right, but the acc_scale is just the same line over. Plus the acc_scale is typically around 1.0 (either less or more), because it factors the measured value and the real value.

MrAhlefeld · March 8, 2021, 10:35pm

I’ve reached out to 2 persons in this thread with tracking issues like mine, for them to test a guide to update their 8KX configurations. This incorporates @Chumet and my work with the 8KX tracking issues.

Once testing is done and if results are as good as mine and Chumet’s, i’ll release it into the Wild here on OpenMR

Br.
Martin

3D_FAN · March 9, 2021, 8:41pm

I would appreciate it. Don´t have time to dig in that deep as You did. Because of the knuckles I do not want to go back to LH 1.0

MrAhlefeld · March 9, 2021, 9:05pm

Found this nifty little tool to check Base Station positions as well

For the following type of Base Stations
Original HTC BS1.0
HTC BS 2.0
Valve BS 2.0

ricknau · March 9, 2021, 10:13pm

I hope that when and if this is all worked out there will be some (fairly) simple well documented calibration process/program provided by Pimax (or whomever). As a high-end retail product (not “enthusiast” model) this is incumbent on Pimax. (Fat chance, right? ) Looking at all the steps you guys are posting so far I cringe to think of myself trying to claw my way through the process. I just want to play…

Chumet · March 10, 2021, 12:09am

@MrAhlefeld, the app “base station FOV viewer” Very useful and simple, you get a very good idea of the useful field covered by the base stations and where they are oriented. It is useful to better focus them towards your play space and see that you have no untracked areas.

On the other hand, I wanted to share that I just gave my 8KX the extreme IPD mod, but that’s another matter. What if I want to comment that when disassembling the casings, I could verify that the sensors, not all but a large part, are fixed to the casing with adhesive tape, and if you touch them you can notice that they do not have a well defined position, even the you can move. I suppose that they are well fixed enough so that they do not fall or move from their position, but that makes me think that in the production process no HMD is going to have the sensors in the same position as the theoretical design, when we look for dispersions. less than 1mm between them. This means that Pimax should be doing an individual calibration for each HMD it manufactures and also the calibration process must be reliable. In a production I do not think they waste too much time in calibrating it very exactly and that they give 100% security that it is well done. And that is what end users are suffering now. And on top of that we have been complaining for months and requesting help, they have not spoken at any time.

I am still investigating the calibration process together with the valve 2.0 base stations. Since my lighthouse 1.0 that I bought are already returned, but I have been sent to calibrate my HMD and obtain a very stable tracking without jitter and I have also verified that the 2.0 are better in many respects than the 1.0. But here I do not want to enter since that comparative has been done previously.

lhoyet · March 11, 2021, 4:34pm

Hi all, I seem to have a similar problem with my 5K+, and even after receiving a replacement HMD yesterday I have the impression that the problem is still there. (similar jittery headset, with Valve 2.0 lighthouses.

I was wondering if someone had a config.json file that works fine for his 5K+, to compare whether the values are the same than on mine or not (and possibly to upload a different json file to see if that improves things or not).

Thanks in advance

MrAhlefeld · March 11, 2021, 6:58pm

Hello lhoyet, is it the “new” 5K+ with the blue coating ?

lhoyet · March 11, 2021, 7:49pm

Exactly. We also have two “old” black 5K+ version (backer’s version), which does not show any of the jitter the new one has.

twack3r · March 13, 2021, 8:45am

How are you getting on with the guide?

I’m in sore need of non-jittery tracking…

Thanks!

MrAhlefeld · March 13, 2021, 10:54am

Just fine nobody has replied with their results.
So I’ll just PM you the details to test.

/MrAhlefeld

MrAhlefeld · March 13, 2021, 10:59am

Hi lhoyet

The placement of the sensors would be the same as the 8KX, so that part of the .JSON would be useable for you.

If @twack3r test and reports back today, I will release the files and description today as well

festr · March 13, 2021, 11:43am

@MrAhlefeld would you send me the procedure to me as well? I have 8kx and would love to test it too (I have jitter too)

twack3r · March 13, 2021, 12:17pm

Just read the instructions, sounds easy enough to do, thanks!

A couple of question before I start:

Find the following line “device_serial_number” : “LHR-ADE69611”, and exchange the LHR-number with your own

How do I find my own LHR-number?

Sub-questions:

Is the LHR number an HMD specific number, or are there e.g. 4 LHR numbers when running 4 basestations?
Uploading the altered .json uploads it onto the HMD? Is that correct?
Would it make sense to pull a back up of the current .json file in case something goes horribly wrong? If so, how would one do that?

And finally:

Due to some yet unknown reason, when running the SteamVR Room Setup the Avarage Rotational Deviation allmost doubles, and the Avarage Positional Deviation is impacted as well.

I don’t understand this part. Are you saying that if running the test whilst running SteamVR Room Setup, the deviations double? If not that, what are you trying to describe?

MrAhlefeld · March 13, 2021, 12:22pm

I think we should continue this in a PM so everyone dose not panic

twack3r · March 13, 2021, 12:27pm

Alright, shoot!

I don’t panic easily

MrAhlefeld · March 13, 2021, 2:53pm

Yes the LHR number is specific for each SteamVR tracked object, controllers, HMD, tracking dongles.
Yes correct, it uploads and stores the altered .json in the HMD.
Sure, just added it to my guide.

And finally
Due to some yet unknown reason, when running the SteamVR Room Setup the Avarage Rotational Deviation allmost doubles, and the Avarage Positional Deviation is impacted as well.

(The above means that when you have initially run Pimax Room Setup, it will have a position for the HMD in the “Universe” but the controllers do not “at least that’s what I have experienced”. Then running the SteamVR Room Setup will make the Chaperone for you playspace, and somehow this definition of the “universe” alters the tracking stability.)

There are 2 different universe files, one in you Steam install dir and one in a Pimax dir.
eg.
C:\Steam\config\chaperone_info.vrchap
C:\ProgramData\pimax\runtime\chaperone_info.vrchap