Comunity powered VR rig?

[virror]

With all the amazingly talented people here at the forums it would be really cool to see what we could achieve if we worked together against a common goal.
That goal would be to make some kind of VR rig with multiple systems working together to create "the ultimate VR experience" : p
First would be some basic spec what kinds of systems would be needed and then narrow down the specs for those individual systems.
Since we have both talanted coders, electrical engineers and mechanical engineers i think we could make some amazing stuff, all would of course be DYI friendly and open source for the best of the VR future : )

Thoughts?

[geekmaster]

... it would be really cool to see what we could achieve if we worked together against a common goal. ... i think we could make some amazing stuff, all would of course be DYI friendly and open source for the best of the VR future : )

Thoughts?

You need common interests to have a common goal. I think having everybody exploring their own path of interest is where the real innovation comes from. Where would the Rift be if Palmer spent his time on some "common goal" instead of persuing his own personal passion of creating the best VR experience ever?

And if you did achieve your goal of making everybody work on a single common project, that would cause problems for nonconformists who choose their own path of exploration. Conformance reduces innovation.

[mahler]

Conformance reduces innovation.

Can't fully agree with that. As always, there is a right balance.

[virror]

I don't agree with that either, there are a lot of great examples that prove you wrong, also VR is something i believe everyone here has as a passion or at least big interest. There are varied enough stuff in this kind of project to allow ppl to work on what the like the best. But sure, some ppl like best to work solo, and some ppl work best together, thats the human nature.
Just curious about what everyone thinks of the general idea.

[geekmaster]

... there are a lot of great examples that prove you wrong ...

Because "innovation" is a derivative of "novelty" or "something new", when everybody is working on the "same old thing" how can there be novelty outside the defined boundaries of the common project? The best discoveries are where people have not been looking, off the beaten path, UNCOMMON...

So, I am not necessarily wrong in my opinion, which remains a valid viewpoint. There is room for both common and independent projects, and you can expect the greatest innovation from independent individuals or teams just doing their own thing, NOT defined by any common goals... IMHO

Different is not wrong. The difference is the novelty...

[MSat]

The VR field is so vast that any goal one would have in mind would in essence be quite specific. I find that my various interests in VR are quit disparate, sometimes with only minimal overlap among them. It would be great to see some of the skilful people here working together, but on what? If you can make a compelling case for something, then go for it! I have sought out interested parties on things I'm working towards. If anyone can come up with something that piques my interest, and I have the necessary skills to help out, then I might be more than willing.

[virror]

Well, the first step would be to figure out what systems we need : p
I posted this mostly as i was curious what the response would sound like, but i will try to make a suggestion.

Since we already have Rift on its way i think a HMD would be pretty meaningless, instead i think a good body tracking system would be a good thing, i suggest a combination of the Kinect and a IMU based system maybe, were the Kinect would do the torso and leg tracking and IMUs would do the arm tracking cause arms are more sensitive to latency and i think Kinect have troubles with them i i stand with my back against it but it can do body and leg tracking very well and a little latency there wont matter as much.

This system could later be expanded with the option of maybe some external I/O card like a Arduino as well to trigger external stimuli.

Some walk in place system would also be cool but i think thats a lot harder to make.

Now time for bed : )

[geekmaster]

Yes, a good affordable, accurate and low-latency body tracker would be a good thing that could benefit all Rift (and other HMD) users. However, not everybody can contribute to it (other than testing it and giving feedback). Some people have skills in areas far removed from sensor technology, kinematics, and haptics. For that matter, some neuroscience may be needed, to help solve "uncanny valley" problems that may arise if the tracking is "almost" perfect. There may be a use for other skills too. Perhaps legal professionals when somebody trips while blinded by VR and significantly damages their real body? And if an ODT is included, that will open up a new set of needs for legal and safety professionals to contribute their advice to the community project.

But just because somebody has the needed skill set does not mean that you can pull them away from their own pet project for very long...

And don't forget, the Rift may be so good as to be totally addicting to some people here, and time inside VR is time NOT spent out here working on the community project... Who will take over their workload when we lose them to VR?

And I am not trying to be a naysayer here... I am all for our contributions to help bring to reality the Holodeck, and the Matrix, and Skynet... Bring it on!

[MSat]

Well, I'm working (more like waiting to hear from Freescale) on an upper body skeletal tracking system, so hey, if someone wants to help out, they're more than welcome

The only thing is I worry that there is no point in the effort as we know tracking tech is a major focus for Oculus going forward, but they're staying mum about it for the time being. If the consumer version is going to incorporate their own tracking system, then what would be the point of our efforts? The safest bet is software related work, but that's not one of my strong suits, so that leaves me out

[virror]

geekmaster: Cant you just keep put of this thread instead of just trolling please?

Msat: What i understood about the consumer version of Oculus is that it would include positional tracking for the head only, and even if they want to do more tracking in the future thats probably years ahead since the rift probably is like one year ahead.

Also to be honest i don't really have a plan with this, it was more like "It would be cool with a community VR platform that made/took some tech and made sure it works together to provide a great experience", just wanted to see what the response would be like : p

[geekmaster]

geekmaster: Cant you just keep put of this thread instead of just trolling please?

Sorry, no trolling intended. I may have misunderstood the intent of your thread. I interpreted it as a desire for all of us to work on a common project, and I was expressing potential complications to take into consideration.

If you interpret my posts as trolling, then your purposes must be different from my interpretation. Perhaps you are just trying to recruit the subset of developers who would be interested in contributing to your project. In that case, you need to propose a project plan with a list of tasks to be assigned to or chosen by interested recruits.

It appears to me from your closing paragraph that you want us to work out a protocol or working guidelines so that our individual projects will work together, which itself is a noble goal, but may be hard to pin down without introducing boundaries or limitations that may interfere with novelty or creative research.

As your thread started out, I thought it was a call to action for all of us to work on your project instead of our own. Sorry for the confusion. Calling me a troll is a bit harsh when I have no malicious intent. If you really are asking me to stop participating in this thread, that sounds like you are not interested in hearing about potential complications of your proposal, which would limit the goals of this thread in an indefinite and unpredictable way.

[virror]

I'm sorry, seems that it was just a misunderstanding, after reading what i wrote again in the first post i kind of understand your response : )
The post was mostly just to see what interest/replies i got as i wrote in another post, i don't plan to try and make the whole community focus on the same goal, that would just be stupid and born to fail. Again, sorry about that, just excited that VR finally is taking off (i hope) and want to try and speed up the process : p

[geekmaster]

I'm sorry, seems that it was just a misunderstanding, after reading what i wrote again in the first post i kind of understand your response : )
The post was mostly just to see what interest/replies i got as i wrote in another post, i don't plan to try and make the whole community focus on the same goal, that would just be stupid and born to fail. Again, sorry about that, just excited that VR finally is taking off (i hope) and want to try and speed up the process : p

Sadly, misunderstanding and miscommunication are all too common in a multicultural medium like this. It is too easy to assume that we share a common context when we post, but others are bound to fill in any missing details or assumptions in different and unexpected ways. And we may also be overly protective of the threads that we create, causing us to take constructive criticism as malicious trolling. We are not alone swimming in this "sea of trolls" called the Internet. There really are a lot of trolls out there, who make themselves feel important by belittling others, but not as many self-intended trolls as it appears.

I have created many threads in other forums that brought the trolls crawling out of the woodwork, especially for my most innovative threads, so I see how easy it is to feel disrespected, even when not intended. Just try to remember that most people view themselves as basically good (even the trolls). Posts that seem malicious or disrespectful rarely had that intent by the poster. After all, they clearly found your thread interesting enough to them to invest time to participate in the discussion, even though it may seem misplaced. That usually just indicates a lack of clarity in the goals of the thread, with no disrespect intended.

I am also among most of us here who are excited about the Renaissance of VR. Palmer Luckey and John Carmack and kickstarter are a big part of this, and will go down in history as the champions of VR if it manages to become self-sustainable this iteration. Truly immersive VR looks promising and affordable at last, but I can only hope to get the direct neural interface I have been waiting for most of my life, soon enough that I still have time to use it. Old age comes sooner than we expect, especially when we have exciting hobbies like this.

Thanks for your post.

[STRZ]

Thoughts?

The ultimate experience due to a community effort won't happen because Oculus (Palmer) monitors this forum and signs everybody with potential including his source code right away

Jokes aside, i'd like to experiment with VR on Linux. A dedicated Linux VR distro would be a cool thing to show off what people are doing.

[mahler]

Thoughts?

I'm interested in this because I also have this vague sense of "The perfect VR setup" - but I'm still struggling with the full implications of what that should be.

Some of the best inventions have been made by accident, but the succes of many succesful projects has largely been determined by a strong vision.
So it's important to get the priorities straight and not try to do everything at once. I know Project Holodeck is already trying to tackle this subject. You may want to have a look at their website.

There are many features to consider

• HMD - What is the resolution you need, FOV, weight
• Tracking - which body parts, how many dof, which precision, which accuracy and when to track (when you are look or not looking)
• Interaction - how do you interact with the world. Speech, touch or movements/gestures.
• Separate peripherals - how many, which kind of feedback
• Environment - how large is the area that you want to support, do you need wireless, thing about influence of bright lights or darkness (or fluctuations)

virror - Can you give us your vision of the perfect Community powered VR rig?

[geekmaster]

... So it's important to get the priorities straight and not try to do everything at once. I know Project Holodeck is already trying to tackle this subject. You may want to have a look at their website.

Project Holodeck is currently based on the Oculus Rift, so it may indeed be the project framework that this thread is looking for:
http://www.projectholodeck.com/system

The current hardware design of the Holodeck system uses the Oculus Rift for head mounted video feedback, the Playstation Move optical system for head tracking, and the Razer Hydra magnetic system for limited body tracking. When combined, these systems allow us to create a realistic 3D space that the user can freely move around in and interact with.

[virror]

mahler: That would be the first thing to clear out : p
I think we should use the Rift as a base and build from there, no use in reinventing the wheel.
I don't think we should aim for wireless to start with as it will introduce more complexity and cost to the project.
I think we should try to support 360 degrees rotation.
I also think that a good starting point will be to focus on the body tracking, like the arms, torso and legs. The best idea i have came up with so far (according to me) is a combo of the Kinect for whole body except arm and IMU based for the arms to get better latency and support arm movement in all directions even when facing from the Kinect, maybe there are better solutions, i dont know. A full IMU solution could be viable but i think Kinect for most parts is good because of ease of setup.

[geekmaster]

... A full IMU solution could be viable but i think Kinect for most parts is good because of ease of setup.

Kinect had a bad reputation for high latency in early versions of the Kinect SDK. In fact, it was bad enough that the Holodeck Project stopped using the Kinect. But in the newest incarnation of the SDK (v1.6), latency is actually pretty good. Just look at the latest "Garry's Mod with Kinect" video for an example of recent progress in Kinect body tracking:

[youtube-hd]http://www.youtube.com/watch?v=cKDRcHab6mc[/youtube-hd]

Notice how well Avatar motion tracks body motion in the lower left corner inset video.

For those of us who have a Kinect, this new lower latency could be a good reason to drag our Kinects back out of the closet.

Of course, the most accurate solution is a fusion algorithm to combine multiple sensors, such as Kinect for body kinematics, and Rift-mounted Leap Motion sensor to give better hand and finger motion when in range, with body-mounted Hydra for better hand coordination when outside Leap Motion range, and perhaps with other sensors as well for more precise positional head tracking. All of these mutli-sensorial data should be fused to give heavier weighting to whichever sensors are in their sweet spot of range and accuracy. And it should perform as well as possible with missing sensors too, so not all sensors are required except when increased accuracy is desired.

[virror]

Looks very nice, latency is actually pretty decent, especially for body and leg movement. But i still think a IMU solution would be the best for arms since those are more latency sensitive, but it would still be possible to just use the Kinect of course, or only the arm sensors. Already ordered an Arduino Leonardo and the sensor boards are almost done to send to the manufacturer : )

Edit: The Holodeck project looks cool, but it looks more advanced than really needed and very cumbersome to set up, its not what i call "consumer Friendly". Im sure its possible to make a lot better system : D

[PalmerTech]

I don't know if a community project really needs to worry about being consumer friendly. By definition, it is likely to be used mostly by people in the community. It gives you a lot more flexibility as well.

If the community came up with an open source VR system that was cheap and consumer friendly, that would be great. If the community came up with a system that had better performance at the expense of being friendly for the average person, that would be even cooler, IMO.

[geekmaster]

Now, while on the subject of Kinect hacks, it is time for some creative Kinect madness:

http://www.youtube.com/watch?v=cVCghLfdzsY

Embedding is disabled for this video, so just click it and watch on YouTube.

The smoky cat outline on the wall makes a nice touch.

Sorry for the diversion, but a good laugh now and then is essential to maintaining sanity.

[geekmaster]

This shows a 3D virtual reconstruction using fused depth and video data from two Kinects:

[youtube]http://www.youtube.com/watch?v=5-w7UXCAUJE[/youtube]

The video is taken from a virtual viewpoint between the two Kinects to the left and right. It would be great to see your own hands in front of you this way while using a Rift.

Adding a Leap Motion sensor to the mix would give much better accuracy for finger positions.

[virror]

Palmer: If its hard to use, than sure its cool but pretty useless, unless some company is interested in it of course : p

geekmaster: Two great videos, first one was very good : D

[PalmerTech]

I would tend to disagree. There is a lot of cool tech out there that is hard to use from a consumer standpoint, but can provide an incredible jump in realism. Galvanic Vestibular Stimulation, for example; You have to attach gel electrodes to your head for each play session, and they can only be reused a limited number of times. Definitely not consumer friendly, but it is something the VR community might put up with. A company, on the other hand, is generally only going to become interested in technology that they can sell to consumers and make a profit on.

Making a DIY, open source "ultimate VR experience" does no good if you rule out hard to use technologies. I would be willing to bet that most of the people here would be fine with spending hours fiddling with drivers and settings, or strapping themselves into a crazy contraption if it gave them a great VR experience.

[mahler]

virror, I get the idea that you are looking for a great VR experience using relatively cheap components with a DIY setup, similar to how the first prototype of the Rift was created. So for example combining readily available hardware like the kinect, leap motion, webcams, wiz dish and mobile phones in a smart way to create a broad VR platform. Although GVS is a relatively untested area in relation to those other technologies and it might be intimidating for anybody who has no experience with it, that could be exactly the type of alternative solution you are looking for.

But as I said earlier, it usually helps to first try and figure out exactly which problems you want to solve... then look for a way to do it. Just the words "Virtual Reality" don't mean the same for everybody.

[geekmaster]

... Although GVS is a relatively untested area in relation to those other technologies and it might be intimidating for anybody who has no experience with it, that could be exactly the type of alternative solution you are looking for. ...

For more information about GVS (links and quotes), read this recent post:
http://www.youtube.com/watch?v=DnXZiEXsSW8

Especially notice the claims that GVS can reduce or eliminate VR motion sickness caused by visual and vestibular mismatches. That alone is good for people who will not tolerate VR queasiness. But for us hard core VR fans, GVS can also increase immersion, which is important to some of us.

I used to play Descent while sitting 5 feet from an 11 foot diagonal projection screen. It filled most of my FoV and was very immersive. It caused horrible vertigo and queasiness at first, but over time I could play longer and longer before the onset of motion sickness. Eventually, I could play for hours. This duplicated my experience learning to read a book in a moving vehicle, which slowly increased from minutes to over an hour (my commute time). Although GVS probably would have helped with the Descent playing, I doubt it would help much with reading in a moving vehicle.

The point is, people who get queasy using the Rift need to either use GVS, or just give it time with increasingly longer sessions, until you no longer get motion sickness while in VR.

[cybereality]

Descent was a serious stomach churner. No doubt.

[geekmaster]

Descent was a serious stomach churner. No doubt.

The biggest gut-wrenching experience was when you turned a corner and your sense of up and down changed, because what was the floor and ceiling are now walls, so you need to rotate the craft.

Astronauts in the space station expressed a similar phenomenon, when their sense of up and down suddenly shifted. Worse, when at the end of a long tunnel in weightlessness, it can suddenly appear that you are at a great height and about to fall, which can trigger panic if you have a fear of high places. Vomiting in weightlessness can be a serious problem from the danger of inhaling vomitus.

How hard would it be to get Descent working with your Vireio driver? I wonder if head tracking would make the motion sickness worse. And I wonder if my adaptation to it years ago has faded from disuse...

However, I want to play Descent in my Rift!

[cybereality]

Someone made a mod for Descent to work on the Rift:
http://www.mtbs3d.com/phpBB/viewtopic.php?f=140&t=15377

[geekmaster]

Someone made a mod for Descent to work on the Rift:
http://www.mtbs3d.com/phpBB/viewtopic.php?f=140&t=15377

Now that's effing awesome! I have not seen that link before. Thanks for the link!

I suppose I need to scan old threads, but I already spend too much time here.