#1
|
|||
|
|||
world is tilting - intersense or code problem?
Have you ever experienced your worlds starting to tilt to the left or right? We are. I am thinking source of the problem is the intersense. The tilt seems to vary over time - possibly randomly. Sometimes it seems to level out. Sometimes it seems to drift.
Any thoughts? I have recently figured out how to make 3d work in our setup, and even more recently had some (limited) success with reseting the intersense. See my recent thread on this subject (I am overdue in updating progress on that thread..I think the thread is called "reset intersense not working"). Perhaps there is something less than optimal about the way I am doing linking to the intersense and/or resetting that is making the intersense confused. You don't have any likely diagnoses bc of that do you? It does not seem to be caused by slipping of the hmd on the wearers head (although there is some slippage the tilting seems more extreme...i think I have also noticed it right at startup when the hmd was level). We thought maybe the Velcro connection btwn intersense and hmd was loose....in the process we actually removed the enter cube from the hmd as it was being worn by someone. Miraculously, the vr scene did not change is orientation no matter how much we tilted the intersense cube! Instead, the vr scene orientation changed only in response to yaw, roll or pitch rotational movement of the head on which the hmd was (still) being worn. How could thus be? Isn't orientation of head measured solely by the intersense cube? Is this behavior anomolous? If so, perhaps it is related to the tilting we are describing in the previous 2 paragraphs? (Note that we do not have our PPT system turned on...the blue ligth on our hmd is on but the cameras and the rest of the ppt software are not running). Any thoughts? One last point....I seem to be getting the same behavior on our other intersense as well. Maybe it is less severe? (Sometimes I doubt my own perceptions of tilt...been wearing the HMD so long that I wonder if I am getting disoriented? That said, other's have definitely noticed the tilt on the first HMD. |
#2
|
|||
|
|||
world is tilting - intersense of code problem
SHORT VERSION:
I seemed to have solved this. The key code change was to remove.... viz.go(viz.QUADBUFFER | viz.FULLSCREEN) viz.link(viz.add('emagine.dls'),viz.MainView) ...and add.... import viztracker viztracker.go(viz.QUADBUFFER | viz.FULLSCREEN) LONG VERSION: I have spoken to others about this. One person suggested I phone emagine. Another suggested that my intersense was damaged (as has happened to that person several times in the past requiring sending it to be repaired/replaced). I spoke to someone at world viz and got some other valuable suggestions: Suggestion 1: do not hold or wear the hmd but rather keep the hmd on a rock steady surface (like a desk) when "starting out". (I wished I remember more precisely what "starting out" refers to...My guess is that "starting out" means somethign like when reseting (via software or the gui in the lab tool) or when powering up or when hitting f5. Can anyone more precisely clarify the critical period(s) when intersense must be steady on a desk or whatnot? This is important as I did not know that! Suggestion 2: if you are using worldviz software hit alt-r when facing z positive direction. Actually none of the above suggestions seem to have directly helped in my case. I post them so that others may benefit. However, Suggestion 2 made me think that I should pay more attention to viztracker.py. Importing that had messed me up (long story) weeks ago in the past when I was much more ignorant than I am now. However, I recently read the code in viztracker.py and noticed that it contained comments about using alt-r for reseting (which is a potential solution to my other thread of a few weeks back called something like "intersense reset not working"). So, I started importing viztracker.py again with the benefit of more knowledge gained over the past few weeks. Like I said at the top of this message, specifically I added this... import viztracker viztracker.go(viz.QUADBUFFER | viz.FULLSCREEN) ...and removed this... viz.go(viz.QUADBUFFER | viz.FULLSCREEN) viz.link(viz.add('emagine.dls'),viz.MainView) ...and the tilt problem seems to have gone away. Also hitting alt-r would successfully reset my intersense. The only minus is that my viewpoint is quiet a bit higher than before. I can fix that will a simple seteyeheight command. Does anyone shed any light on why... viz.go(viz.QUADBUFFER | viz.FULLSCREEN) viz.link(viz.add('emagine.dls'),viz.MainView) ...seems to cause my horizon to tilt? Well I have raised a large number of questions on this thread but my urgent issue has been resolved. I would still like to get the above questions answered so that I have more of clue what I am doing! (-: (If it becomes too complex to track all the questions we can make new threads for each one) |
#3
|
|||
|
|||
It looks like you were using the sensor that is built into the eMagin visor, and not the intersense cube here:
viz.go(viz.QUADBUFFER | viz.FULLSCREEN) viz.link(viz.add('emagine.dls'),viz.MainView) because you are using 'emagin.dls' rather than 'intersense.dls.' That would also explain why you had no control over the viewpoint when you rotated the cube, nor when you tried to reset it. But when you used this: import viztracker viztracker.go(viz.QUADBUFFER | viz.FULLSCREEN) the viztracker module detected your intersense cube, and things behaved as expected. That's my guess anyhow. We also use an eMagin visor, and as I recall, it links by default to the mouse cursor, so that might explain your viewpoint's odd behavior. You can turn off the eMagin sensor by opening the Visor Utility and checking the "Disable Sensor" box on the Motion Sensor tab. I hope that is useful. Good luck.
__________________
Virtual Environments and Multimodal Interaction (VEMI) Lab This time, it should work... |
Thread Tools | |
Display Modes | Rate This Thread |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
reset intersense not working | billjarrold | Vizard | 9 | 03-31-2010 10:53 PM |
intersense with 5DT glove | sjp | Vizard | 2 | 04-06-2009 08:22 AM |
intersense orientation | jargon | Vizard | 2 | 05-14-2005 05:55 PM |
Problem with newer Intersense Inertiacube 2 | esmith4 | Vizard | 3 | 03-18-2005 09:13 AM |
Vizard Crashes: causes are hard to determine, possible problem with the viz code | vr_boyko | Vizard | 1 | 01-07-2005 11:52 AM |