"World moving with you" phenomena. What, Why and Fixable?

Pimax 8K & 5K+ Full Review | The Most In-Depth Pimax 8K vs 5K+ Review and Analysis you will find! - YouTube headset wont even power up without external power coming into splitter/breakout box.

1 Like

Ok, thanks for that link. So we can assume it is not a lack of power or power being redirected.

you are making things up here without any good reasoning
Occam’s razor
it was already narrowed down to scaler or mipi/display and the scaler seemed a more logical way as it can more or less easyly overclocked and tested how fast (picures per second) it can do
just becuase the prototype in vegas had tracking problems because of additional led’s does not mean its related to another (much older already settled problem)

1 Like

I added question marks in there. As I wasn’t sure!

Ok, so the scaler got to 85Hz using overclocking yes? But that was not stable so they dropped it to 80Hz. So what are the parameters of the scaler that can be changed to allow that overclock back without the stability issues?

What can be done in your opinion?

Stupid is being a little too wound up about whats written in Internet forums.

1 Like

Ok, but if tracking and power are somehow at their threshold then the post by @LoneTech above that says it might also be tracking related brings us back to power?

I don’t think there’s any power issue (but obviously I haven’t measured anything). What I was talking about is the motion to photon model versus reality; tracking always predicts where you’ll be when the frame shows up. If that prediction is a little earlier than reality, the world will move along with you. This time difference is dependent on the frame rate, so if switching frame rate didn’t switch prediction offset, it explains the phenomenon. And with Pimax working on more frame rates, they’d encounter and fix that sort of issue.

1 Like

They did play around a lot with the 8K panel Hz. But something like that would be obvious in Pimax internal testing I would have thought. I can’t find it now but one of the testers said the world moves with you in the peripheral, it might have been a reddit comment which I am trying to find and hope they were not mistaking it with the wide fov distortion.

I haven’t located the comment about it being in the Peripheral yet but @knob2001 in his review said it is in front of you. If this helps.

The aberration is a mixture of both, the barrel and pincushion that feels like a “mostacho” effect. It is evident? yes, it is evident at Large and Medium. On small, the black blinders hide most of it. But the problem, the biggest problem of all, is that the aberration along the whole vision, although not very noticeable, creates a weird feeling that the world in front of you is following your movement. Sebastian talked about this. You probably have heard about it too.

the outer area was not what mrtv’s complains was about
we know about the distortions in the outer area and because of the not correctly handled positions of objects they are misplaced and look weird (missing compensation for lens distortion)
for now that is the reason to have large/normal/small like 170°/150°/120° as horizontal field of view (just dont render this area, saves gpu power and reduces perceived distortion in peripheral area)

the thing we are about here is whats directly in front of you in a optimal area where math (movement compensation by sensors and distortion correction for the lenses) should guaranty a good illusion, like it is the same as in the real world

the 8k 80hz ship has sailed a long time ago, the unit is in production now, don’t expect it to be pushed to 85 or higher, the reason for that was not revealed by pimax we have to live with assumptions and that they said that there is a limit in the hardware they can not solve, that was on debate and testing starting from november 2017 and pimax did not manage to change it (for a reasonable timeframe and price) so it stays that way (as announced in the kickstarter before the end - anyone with doubts had the option to cancel before the end)

Yes I know.

Yeah, I hope a fix can be found, a software fix ideally. @LoneTech has put forward some decent info on this subject.

The thing is that, if you’re talking about the real o virtual review and patreon tests, they tested the 5k, which supposedly doesn’t have that issue since it’s 90hz, so I’m really confused about this.

1 Like

Yeah, It could be a possibility that all these testing headsets are different qualities which is causing a lot of debate why one person can see something another can’t. Also, LoneTech has mentioned it could be tracking. Which might be a better thing as the Hz issue would not be important (again).

Correct me if I am wrong, or any of this is outdated.

Sweviver can see a black spot pattern on orange and skin tones
Sebastian can not see this.
Sebastian said the world moves with you.
Sweviver said this was fixed in a PiTool update.
Silence…
Knob2001 says the world in front of you is following your movement. (they got a late 5K+ review / test unit)
Sebastian can see the world moving with you again (8K), described recently as a wobble effect

Or maybe that refresh rate is not the real reason behind that issue, it’s still a theory after all and Pimax told Seb it had nothing to do with that.

2 Likes

im not on the level as @LoneTech but the bets are not off that it can be fixed by a better lens profile, if people perceive distortion then it sound like a lens problem

in the vive there are correction values in the firmware of the headset, they are measured in the factory for every lens and are different for every lens, with the gearvr mod you had to read that values before altering them or you would be screwed when returning them, one of the problems with this mod was that the distortion profiles for the gearvr is stored online (serial number) and when putting these lenses in the vive they where completely ignored
what people found out is that you might see distortions and depending on how much off the lenses where and how perceptive the person was it came out from best,over ok to total sh…, you could lower that by manually altering values and testing, that is extremely time consuming and does not guaranty a optimal result, the only solution would be to measure the lens as be done in the factory and create a optimal lens profile and i guess even then you there would be people seeing problems as the whole thing with this lens type (non fresnel, just pancake) its seems to be more dependent of eye position, eye shape, …? there where reasons why valve decided for fresnel lenses even as the hat not that good sweetspot

so what i was about with this vive/gearvr story is that lenses can be different and atm we dont know how pimax handles this, maybe there are some varieties in lenses that got ignored or mesured wrong and testers can see this, depending on how bad it actually is and how perceptive the person is
it can be pretty hard to drill down on that without having access to the hmd, i guess the tech people at pimax would like to have all units in question in the lab to compare the lenses and profiles but that’s not a option
also mounting problems could occur, remember, the production line is new and the hmd’s out there are something between prototyp’s and 1st production run

would be interesting to know how pimax handles the variety of the lenses and how this is stored/used

3 Likes

I am surprised that HTC do per lens profiles as that is a costly human process. It is how DSLR lenses get chosen during manufacturing and why the best lenses end up in higher end products for their focal quality. But these are cheap $30? lenses and I assumed they all have the same profile and a simple pattern test would be either a pass or a fail. Just an opinion.

I agree it would be very interesting to know if lenses and distortion profiles are matched together on a per lens basis.

Still would not explain why SweViver sees a black spot pattern on skin tones and MRTV does not.

Also, would this create a wobble effect if the lens profile is not matched well? Sounds like it would just give a poor image.

Slight lens aberrations should cause neither wobble nor the faint dot pattern (they’re not black spots, just off-colour patterned dots).

It would however explain edge distortion differences between headsets being ‘barely there’ to ‘noticeable’ to ‘unusable’

1 Like

Was it StarVR that said they match each lens to a distortion profile and why their headsets will be very high end. Which to me is another way of saying it will cost a lot because of that effort. Is it reasonable to think Pimax will be doing this?

No idea, they’ve never mentioned it either way. We don’t know much about their manufacturing steps or quality control processes.

When was the wobble, world moving with you introduced? From the start or later?

it could be possible that Pimax have spent so much effort trying to rid the lens distortion that the current algorithm (or a bespoke version for MRTV’s lens) takes too long, and now there is latency causing the wobble.

It would be interesting to know if the wobble gets worse on lower end kit.