Taken from the Safety specs on the two.
HTC v1.0
This product contains Class 3B laser, which can produce hazardous level
of Class 3B and 4 laser radiations.
SteamVR v2.0
The SteamVR⢠Tracking 2.0 Base Station contains a Class 3B laser, which can produce hazardous levels of laser radiation.
But thatâs just the BT radio I believe, or is it 
I really donât know for sure.
Anyone care to chime in who knows ?
There are more factors in play. While the âclearâ 5k+ looks impressive, I believe it may also have an adverse effect on the tracking if the shell starts acting as a light âconductorâ. I guess it does not, but I could imagine that when the laser hits close to the sensor spot the inducted illumination may transmit through the shell and disturb the sensor at least.
The blue coating may work as an attenuator (at least it does when shot with your camera), and therefore actually mitigate the effect I described above. On the other hand it may as well attenuate too much making it harder for the laser to âmake it throughâ especially at steep angles.
This is just a speculation though, but in general, I guess your observation is pretty much in line with the theory that the coating did more harm than good.
I am not sure if HTC Vive (and Pro) are using non-transparent plastic (but the way it is molded suggests they do), but the black circle is a thin IR transparent cover which protects the sensors. This way they eliminate any induced light into shell and also clearly define the angles at which the sensors can activate. Compared to Pimax where with the fully transparent shell, the sensor can be hit basically from any direction, some of which probably you would not want.
It is BT radio. It is written in the document.
The covers on the vive would be akin to the lens cover on a remote.
This has given me an idea and instead of removing the blue coating, which is really resistant, of higher quality than expected, since I tried scratching a small area on the inside where the conford kit is attached, and I I realized that it is not an option. Today I have covered all the sensors with a sheet to darken car windows as if putting a filter. And the jitter has improved a lot, now everything is much more stable, the values ââdo not exceed me in âPosition Standard Deviationâ from 0.5 in the worst case to 0.2 in the best. And in rotation they go from 0.2 to 0.1. And I can assure you that it was almost always above 1 in position and 0.5 in rotation. All this has been a quick test that confirms that the jitter is given by a bad coating of the sensors where the blue paint penalizes but does not obstruct the passage of infrared rays from the Valve 2.0 base stations. I am sure that I can further optimize the sensor filter with a sheet of more specific characteristics for infrared filters. What I have used is for critical UVA rays and its origin is from the Chinese bazaar on the corner of the house for ⏠2 cost. I think we are going in the right direction with the subject and at the moment I notice that I have improved the jitter in PIMAX 8KX with blue case, of course.
If someone more expert in these matters of filters or lighting can propose what material would be suitable, it would be of great help.
Some photos from my quick test:
Who would have thought that REDUCING the amount of IR signal would improve things?
Hold on though - these are not good results. Quite bad, actually - good tracking starts below 0.6 max position deviation. What makes you think this was an improvement?
Did you do before/after tests?
This has made me have less jitter and lower jitter. The measurements at different positions of the HMD are also more even. My impression is that it gives more stability to the follow-up. I know my data is still bad, but I think due to my tight environment and valve 2.0 base stations I am not going to get better results. But Iâm going to keep trying because if I notice that there is a greater distance, less jitter, since I cannot separate the bases further, I have set them as high as possible 2.7 meters. So putting a filter on the sensors makes the effect that the IR signal arrives with less intensity, perhaps simulating more distance. But I think Iâll end up buying a couple of 1.0 base stations.
Seriously? Do you have a better alternative?
To date, Valveâs Lighthouse system is the top tracking system for VR. admired by all.
Go watch Alan Yatesâ presentation on it to appreciate the complexity.
EDIT: Oops, for some reason this post was shown as the latest reply to the thread. Anyway, my response is still valid.
Again, before vs after numbers?
Iâm getting seriously annoyed by people posting snakeoil solutions triumphantly, without any facts to back them up.
You canât be serious.
Here is the post of earlier. Before it was jitter 0.3 to 1. Now max 0.5 with low of 0.2
https://community.openmr.ai/t/who-has-stable-lh-2-0-tracking-on-8kx/31984/297?u=heliosurge
Sorry, those are still some random numbers thrown in, not even specifying which measurement heâs talking about.
Not exactly before - after.
And the âafterâ numbers are simply bad. So no way to tell what this hack achieved, if anything at all.
I remain unconvinced.
Only way to see would be to try the film used as he did detail it in his earlier post. He did say a different film/filter may yield better results.
As with any idea it would need multiple tests also conducted by more than one tester.
His range of numbers does suggest improvement; even though not stellar.
I have repeated the test without foil, with a single layer and with a double layer of dark foil. The difference is really little, at least at that exact point where I have positioned the HMD and I have even marked the position to repeat the tests in exactly the same place.
Results:
Without dark foil:
Max Deviation 1.78
Standard Deviation 0.45
Rotation 0.3
With dark foil:
Max Deviation 1.73
Standard Deviation 0.35
Rotation 0.18
With double dark foil:
Max Deviation 1.56
Standard Deviation 0.34
Rotation 0.23
On the other hand, I have been able to take photos of the sensors with an infrared light flashlight and the mobile camera. And there is total transparency of the case and the foil. I have also picked up some Knuckles sensors and they look very similar, except that the housing is opaque and there is only a small circle uncovering the sensor.
So I come to the conclusion that it is not the transparent housing or the blue paint that is the culprit, and that it does not greatly improve putting fancy filters to the sensors, but it does help stabilize the jitter.
Iâm afraid the way it interprets the signals from the valve base stations 2.0 should be optimized to filter out jitter. Perhaps it is to interpret the signals of the 1.0 base stations, which I understand have another way of transmitting the positioning and synchronization between them.
And by the way, when removing the pieces of black sheet, the blue coating has been tinted and there is no way to remove it, at least with alcohol. What a grace, but it does not affect the bad monitoring that I already have. For the moment I will continue with the double dark foil and the protective silicone sleeve.
Could you explain how you have updated the firmware from 3002 to 3008? Without an HMD from HTC or INDEX ?. I have a lighthouse with firmware 3002 and the other 3008. I know there is no difference in monitoring but I like to put it to sleep instead of standby with the android apk âLighthouse PMâ. I have tried to update the firmware but have not been successful, I even bought the linkbox Pro to have bluetooth support in steamVR and it still did not offer the update and it showed âbluetooth service not availableâ although it was visible in the device manager. Copying the lighthouse 3008 file from the âCRP DISABLDâ folder and pasting it into the other lighthouse did nothing either. And on top of that the power to turn it off remotely broke. When copying the original firmware file of that lighthouse again, everything was normalized. But Iâm still on version 3002.
How did you get it? help with this, please.
Thank you!
Now THIS is what I call a well documented experiment. Well done.
The only thing is, these numbers seem to vary between measurements without changing anything, so Iâm afraid the variance is within margin of errors. I could produce results similar to your numbers simply by pressing space repeatedly.
So unfortunately I agree with you, the foil doesnât seem to do much to improve the tracking.
The blue coating is probably a TPE (https://www.reddit.com/r/hardware/comments/6fznfp/cant_companies_stop_using_coatings_that_make/) which gets sticky and eventually completely melts when exposed to the oil or other human skin by-products. There is no way to clean it with alcohol, or any other dissolvant (you would only make it worse).
I have an idea. One of the Key differences between say the 5k+ and the 8kX is the power delivery. With earlier headsets it had a separate 12v wall adapter; where as the 8kX has dual Usb.
Is it possible that the sensors are not receiving the same amount of power in terms of Amperage in the 8kX vs the earlier models?
I recall was it the v4 prototype had power delivery issues when they added another strip of leds to the backlight which caused tracking issues?
This could be why v1 LHes are working better due to more powerful lasers maybe.
One idea might be to use a usb splitter to have more power available as some reported of usb power over current on motherboards.
A test to see power draw affecting tracking would be to plug a device into the lower usb-c port and see if tracking is worst.
