POSITTRON: yet another proposal for positional head-tracking

This is for discussion and development of non-commercial open source VR/AR projects (e.g. Kickstarter applicable, etc). Contact MTBS admins at customerservice@mtbs3d.com if you are unsure if your efforts qualify.
druidsbane
Binocular Vision CONFIRMED!
Posts: 237
Joined: Thu Jun 07, 2012 8:40 am
Location: New York
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by druidsbane »

Not that you need to hear it anymore from this thread but great work! Really excited to see where this goes once the Rift prototype is out.

So based on these numbers the latency is low enough that we can get true absolute positioning with the Rift without added latency over the regular orientation tracking? Really nice work, and it would be great to see how this fits in with the Oculus SDK, if you can write a driver for example so the SDK returns positional information properly (another post suggests that it is supported just won't return anything at launch until there is hardware to slot in there).
Ibex 3D VR Desktop for the Oculus Rift: http://hwahba.com/ibex - https://bitbucket.org/druidsbane/ibex
MSat
Golden Eyed Wiseman! (or woman!)
Posts: 1329
Joined: Fri Jun 08, 2012 8:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by MSat »

I've been keeping my eye on this thread, and my "Great work!" comment is long overdue :) Your approach is very slick!

As a thought, if you want to make this available to DIYers that are weary of soldering, perhaps these people could leverage the huge availability of various battery powered LED gizmos on the market. For instance, there might be some usable products from websites such as:
http://www.raveworx.com/
http://www.flashingblinkylights.com/index.php
ebay
etc.

Another option for those adverse to soldering is crimping, or using suitable jumper wires (http://www.ebay.com/sch/i.html?_trksid= ... s&_sacat=0) and a little bit of glue. With clear instructions, it should be easy enough for just about anyone.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

druidsbane wrote:So based on these numbers the latency is low enough that we can get true absolute positioning with the Rift without added latency over the regular orientation tracking?
Well I won't know for sure until i receive my Rift developer kit and start messing around with the adjacent tracker and some sensor fusion, but i am quite optimistic on this front. Orientation tracking shouldn't be any worse than it is without adding PosiTTron. Worst case scenario (most likely) translations won't be as smooth as rotations, but still usable. Best case, they will be as good and you won't be able to tell the difference.


@MSat

Thanks for the comments!

Yeah i think no-soldering required would be a big bonus for the project. However it was already hard enough to find the individual LEDs with the right characteristics, viewing angle, brightness, etc and then adjust the precise amount of current that goes though each type of LED for all to work. I have my doubts that we could find a LED gizmo that would fit into the design, but you never know! That gave me an idea though, if the positional tracking thing doesn't quite work, maybe we can still sell it on that raveworx.com website.. with the PosiTTron on your head i'm sure you would be the king of the rave party! (especially useful if you're a bit ugly like me) :-P

But yes, crimping and jumper wires could be a good alternative to soldering!


Actually since i think i got some positive feedback and general interest in all this, if i go the open-source route maybe i'll ask for a bit of help to any of you DIY pros watching, someone interested in designing and testing the actual final hardware specs. My electronic/mechanical engineering skills are not particularly amazing, and also my GF is starting to get a bit tired of me spending so much time on this project! (and she's got a good point actually). I would still have my hands full with the whole software side, which still needs a lot of polishing..
MSat
Golden Eyed Wiseman! (or woman!)
Posts: 1329
Joined: Fri Jun 08, 2012 8:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by MSat »

PatimPatam wrote:

Yeah i think no-soldering required would be a big bonus for the project. However it was already hard enough to find the individual LEDs with the right characteristics, viewing angle, brightness, etc and then adjust the precise amount of current that goes though each type of LED for all to work. I have my doubts that we could find a LED gizmo that would fit into the design, but you never know! That gave me an idea though, if the positional tracking thing doesn't quite work, maybe we can still sell it on that raveworx.com website.. with the PosiTTron on your head i'm sure you would be the king of the rave party! (especially useful if you're a bit ugly like me) :-P
Yeah, I figured finding appropriate devices would be a bit of a challenge, and maybe near impossible. It's because of that, that I started thinking of other non-soldering options. BTW, I was reluctant to post that rave junk website, but it was one of the first google results. Still, the PosiTTron would definitely be the ULTIMATE RAVE GEAR! :lol:

PatimPatam wrote:
Actually since i think i got some positive feedback and general interest in all this, if i go the open-source route maybe i'll ask for a bit of help to any of you DIY pros watching, someone interested in designing and testing the actual final hardware specs. My electronic/mechanical engineering skills are not particularly amazing, and also my GF is starting to get a bit tired of me spending so much time on this project! (and she's got a good point actually). I would still have my hands full with the whole software side, which still needs a lot of polishing..
Your skills seem more than adequate, and your renders are great! I'll definitely want to try this out when I get my Rift, and hopefully it can be incorporated to work in some games.

It would certainly be nice if all the correct components could be purchased from just a few locations. Hopefully some of us can help generate a list.

I was wondering about your use of Gaussian blur. If the light sources themselves were diffused like they are on the PS Move controllers, do you think it would still be necessary?
German
Binocular Vision CONFIRMED!
Posts: 209
Joined: Tue Nov 23, 2010 5:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by German »

PatimPatam wrote: The lights don't really need to be dots but they would have to have a simple shape which is easy for the blob detection algorithm to find where it starts and where it ends. Simplifying a bit they couldn't have "holes", so complex symbols like letters wouldn't work very well. But triangles, squares, pentagons etc should be fine. Actually rectangles is a particular case that could be quite interesting.. if you take it to the extreme you could create some cool "paths of light".
If you went with shapes instead of lights, couldn't you make IR reflector shapes and then have a system like TrackIR where there is an IR emitter at the camera? Then you don't need any colors and you can use a visible light filter on your camera(easy mod).
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

@MSat

Yeah i'm quite happy with how the renders turned out, but modeling something that is actually 3d-printable, that would fit properly and it's strong enough, with all the well-thought cable paths, mounting holes, etc, will still take a considerable amount of time and effort i think.

As i mentioned on my last update i'm not using a gaussian blur anymore, but the standard blur still takes quite a few ms anyway. I'm already using diffused LEDs, maybe an extra diffusion layer would help.. other things i want to try include playing with the camera focus and trying to put different types of filter in front of the camera lens.


@German

Yes that's a good idea actually, but then you would have the new entire problem of recognizing different shapes from any possible angle. Not impossible, but that would probably take more precious computing power (milliseconds) than simple color blob detection.
MSat
Golden Eyed Wiseman! (or woman!)
Posts: 1329
Joined: Fri Jun 08, 2012 8:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by MSat »

For a DIY unit, it might be better to try to avoid requiring any sort of 3D printing (unless you wish to sell them, of course). Maybe making an enclosure from foam-core boards that can be slipped over the Rift's housing would be more suitable?


Using reflectors seems like a pretty interesting idea, as reflector tape is available in a variety of colours (http://www.findtape.com/category/reflec ... Qm2hqWTrmI). Then you can perhaps place either some "white" LEDs, or a multi-colour LED array next to the camera. Not sure if you could use some kind of diffuser over them (or if it would even be necessary) though. If such a thing could work, it would make building the posiTTron a whole lot easier!
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by Fredz »

Ball reflectors were used in the past for head tracking with shutter glasses, like with the older StereoGraphics CrystalEyes :
http://www.abs-tech.com/produto.php?id_produto=918

Image

Here is another project using them also for Wiimotes : http://idav.ucdavis.edu/~okreylos/ResDe ... tures.html

It's a lot simpler and more cost effective than using LEDs but you need an IR filter on the camera as it was said and you also need more control on the lighting in the room, since any reflecting surface could be a false candidate for tracking.

Using balls should also probably simplify the recognition algorithm I guess, but you need to use different distances between the balls to correctly identify their corresponding positions.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

As i mentioned on the intro that type of setup with sticking out antennas / balls is something i wanted to avoid if possible. There are already commercial solutions for tracking something like that with specialized cameras and hardware (which cost an arm and a leg btw), but it's really a complete different approach to the problem..

The thing is with my design i need to be able to differentiate at least 4 different types of markers, and i don't think any available off-the-shelve IR solutions allow you to do that (i did look into it when i was researching this).

If you stay on the visible light spectrum i think normal color reflectors would be even more dependent on lighting conditions and viewing angles.. trust me, it's already hard enough to reliably detect and discriminate 4 types of constant diffuse LED lights. Maybe it's worth another try but don't hold your breath on this!
MSat
Golden Eyed Wiseman! (or woman!)
Posts: 1329
Joined: Fri Jun 08, 2012 8:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by MSat »

I admit, I didn't give much thought to my comment. I have no idea how wide the reflective angle is for various tapes, but I'm guessing not very. Even if it is wide, the reflective "intensity" (for lack of a better word) probably isn't very consistent as the angle changes. It would certainly be nice if it could work, but I won't hold my breath either.
Scog
One Eyed Hopeful
Posts: 27
Joined: Sat Nov 10, 2012 4:27 am

Re: POSITTRON: yet another proposal for positional head-trac

Post by Scog »

This brings to question whether the variance in brightness of the reflectors could give useful information to the fusion algorithm.
I imagine there would be reflective tape that might peak front on, or at the angle between the camera and an illuminating light source. It's also likely that such tape could have a predictable falloff corresponding to the angle at which it is viewed, one that could be modeled based on the main tracking data and compared to the video feed, potentially improving tracking accuracy for some axis.

Anisotropic 'brushed' reflective tape could be used, reflective in visible or infrared spectrums, to give strips of distinct axis. it would still need to be diffused to spread the reflected highlight over a sufficient viewing angle. Something like 'brushed satin' would be ideal.

I am also wondering if we can tackle the feature labeling problem with a pattern matching algorithm to eliminate the need for multiple colours. Has anyone looked into what would be involved with that kind of approach? I think I need to go back and read the thread again...
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Some interesting ideas in there Scog!

About reflective tape i think one of the main problems would be that, as you said, the brightness would depend on the position (and type/intensity) of the environment light sources, and i would like to avoid that if possible..

About the pattern matching i think it would be quite difficult to have a reliable algorithm that could do that at 60 fps, with variable angles, from a considerable distance (2 - 3 meters) and with the low resolution of a PS eye camera. Some mobile phone apps i've seen struggle to do that from a very close distance (few inches) and a HD camera..

For now i want to focus on making the current prototype more stable, increase detection distance and lower the latency, without any major changes to the design (unless there's something really obvious that i have missed of course!).
Scog
One Eyed Hopeful
Posts: 27
Joined: Sat Nov 10, 2012 4:27 am

Re: POSITTRON: yet another proposal for positional head-trac

Post by Scog »

While I agree it is probably best to focus more on refining your current method, I think it is worth exploring complimentary ideas even if they appear computationally expensive at this time. I think the variable lighting and calibration issues that come with that may be possible to sidestep by doing (near) real time calibration to match the data curve to that of the primary sensor data, and then use the comparison to reduce noise in the primary data? I don't even know how much noise there is to get rid of, or what kind of data resolution this method could provide but I thought it would be worth asking about.

With regards to pattern recognition for point labeling, is this something that might be within the abilities of modern neural processing methods? Once again this is something that may seem computationally expensive now, but I think it won't be long before it is doable with common hardware, and it could provide for simpler (single infrared wavelength vs multiple visible leds) hardware and potentially more robust tracking in the future.

Just some ideas to throw into the pool of consideration, I am very impressed with your current solution and look forward to using it as soon as possible. :)
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Thanks for the input Scog (and everyone else), even though it may seem sometimes that i'm simply dismissing them, i really appreciate all the new ideas!

About the reflective tape my feeling is that even with calibration it would still be problematic.. on the other hand neural networks have been proven to be able to help on pattern recognition, but i would have to investigate further if they would be useful in this scenario. In general i prefer to avoid probabilistic approaches though..

The main problem as always is time; even if i had a full-time job working exclusively on this I probably still wouldn't be able to test everything i would like to try.. Unfortunately at the moment i only have 4 or 5 hours a week to work on this project (if that) so i have to be extremely selective!
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

One thing i'm slightly worried about, where some advice would be really appreciated, is the legal / patents side of things.. where i must admit i don't have a clue!

The thing is i believe i've seen much sillier ideas than this being patented. Unfortunately i don't really have the time or money to apply for a proper patent (i read the whole process of an international one could go up to 10000 euros easily), apart from the fact that i'm generally against the whole concept (although i understand companies using them to protect their IP against other predators).

I'm simply afraid that someone might take these ideas from here, file a patent and then tell us that we can't use them because we are infringing that patent. What i would like to know is: is the fact i have published this publicly on the internet enough to prevent anyone obtaining a patent? Shall i formally make the HW "open-source" somehow? Or shall i try to print this research on a magazine? (university magazine would work?).

From:
http://inventors.about.com/od/patents/a ... ovelty.htm

"An invention cannot be patented if: 1.The invention was known or used by others in the United States, or patented or described in a printed publication in the United States or a foreign country, before the current applicant filed for his or her patent. 2.The invention was patented or described in a printed publication in this or a foreign country or in public use or on sale in this country more than one year prior to the application for patent in the United States"

I don't like the part where it says "printed publication".. but then again what does "known or used by others" mean? Does this very same post mean it is "known by others" now? Should i be worried? Anything i can do?
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by Fredz »

I'll never understand this fear about patents, most of them simply depict obvious techniques and wouldn't stand in a court.

In this particular case, optical tracking has been used for at least 40 years in the VR field, it's certainly nothing new and the technique you use can certainly be considered as obvious.

Example of prior art :
An early example of an optical system for tracking or motion capture is the Twinkle Box by Burton (Burton, 1973; Burton & Sutherland, 1974). This system measured the positions of user-worn flashing lights with optical sensors mounted in the environment behind rotating slotted disks.

Source : http://www.cs.unc.edu/~welch/media/pdf/ ... h_etal.pdf
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Thanks Fredz for the reply!

Of course the general idea of optical tracking is nothing new.. and while I completely agree that these techniques are quite obvious and shouldn't be patentable per se, after seeing things like Sony patenting a controller that splits in two, or Apple patenting what is basically ambilight for HMDs it makes you wonder.. The fact is i haven't seen any other tracking systems using the truncated cuboid principle to ensure non-occlusion, or a combination of colored LEDs and pattern recognition for marker identification, to name some..

Anyway I don't really want to get into the debate on whether these ideas are patentable or not.. i just would like to know: supposing they were, would they still be after being published on the internet? Does anyone know? Just curiosity really!
Last edited by PatimPatam on Mon Feb 04, 2013 6:04 pm, edited 1 time in total.
WiredEarp
Golden Eyed Wiseman! (or woman!)
Posts: 1498
Joined: Fri Jul 08, 2011 11:47 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by WiredEarp »

I'll never understand this fear about patents, most of them simply depict obvious techniques and wouldn't stand in a court.
I think that fear is when you don't have the money to actually fight and win a court case. While PatimPatam's techniques here might not be an issue, I do think he has some good questions regarding what can be patented and when. After the FOV2GO thing, its something we should all be aware of.
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by Fredz »

You mean the patent brought by the TOYin3D guys in the FOV2GO thread ?
User avatar
mahler
Sharp Eyed Eagle!
Posts: 401
Joined: Tue Aug 21, 2012 6:51 am

Re: POSITTRON: yet another proposal for positional head-trac

Post by mahler »

PatimPatam wrote:Anyway I don't really want to get into the debate on whether these ideas are patentable or not.. i just would like to know: supposing they were, would they still be after being published on the internet? Does anyone know? Just curiosity really!
First to file versus first to invent

[In Europe] When two people apply for a patent on the same invention, the first person to have filed his application will get the patent (assuming the invention is patentable, of course). This holds even if the second person did in fact come up with the invention first. The only thing that counts is the filing date.

In the USA, a slightly different approach is used. In case of two applications for the same invention (a so-called interference), a determination is made who invented it first. This usually involves examining laboratory logbooks, establishing dates for prototypes, and so on. If the person who filed later is found to have invented earlier, he may be awarded the patent.
Grace period

[In Europe] If the invention has become publicly available in any way before the patent application was filed, the application will be rejected (Article 54 EPC). "Publicly available" includes selling the invention, giving a lecture about it, showing it to an investor without a non-disclosure agreement (NDA), publishing it in a magazine, and so on. It does not make a difference whether the person making it publicly available is the inventor, one of the inventors, or an independent third party.

The USA has a one-year grace period (35 US Code section 102). This means that the inventor can freely publish his invention without losing patent rights. However, this only applies for the USA. If an inventor does so, he automatically loses all potential patent rights in Europe (as well as many other countries in the world).
Differences between US and European patents
http://www.iusmentis.com/patents/uspto-epodiff/

More about "First to file versus first to invent"
http://en.wikipedia.org/wiki/First_to_f ... _to_invent

Note that "... United States, which will switch to a first-to-file system on March 16, 2013 after the enactment of the America Invents Act."
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Thank you mahler for the info, especially for pointing out the difference between Europe and US patents.

Still not crystal clear if posting the innovation on an internet forum is legally considered making it "publicly available" but i will assume it is (more so than any other method imho). Also interesting is the "grace period" part, which i understand protects the original inventor..

Ok i think i have a better understanding now, i guess we can move on to more interesting stuff! :-P
WiredEarp
Golden Eyed Wiseman! (or woman!)
Posts: 1498
Joined: Fri Jul 08, 2011 11:47 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by WiredEarp »

You mean the patent brought by the TOYin3D guys in the FOV2GO thread ?
Yep, that was what I was referring to. If they can get a patent on an obvious thing like that (which is probably just a modern rendition of stuff done long ago), then I don't think its unreasonable to at least question whether people could be at risk by disclosing their ideas or products...
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by Fredz »

The problem is anyone can obtain a patent on obvious things since there is no verification by the US government. But that doesn't mean it would stand in a court, and the TOYin3D guys certainly know that since there doesn't seem to have been any follow-up to this ridiculous affair.

It's true that the rules have changed recently concerning the US patent system, but if my interpretation is correct the best thing to do would be to talk about the "invention" as soon as possible so that none can patent it.

See : http://en.wikipedia.org/wiki/First_to_f ... _to_invent

Especially this part (and most importantly the last line) :
Assume Tom conceives of a new mousetrap on January 1, 2006. Tom works diligently from January 1, 2006, to February 1, 2006, to prepare a patent application, and Tom files his patent application on February 1, 2006. Thus, Tom constructively reduced his invention to practice on February 1, 2006. Assume Jerry conceives of the same mousetrap on January 10, 2006, and diligently files a patent application on the new mousetrap on January 20, 2006. Under the first-to-invent system, Tom is entitled to the patent on the mousetrap, because he conceived the mousetrap before Jerry and still worked diligently to reduce it to practice by filing.

As a further extension of the example, assume Tweety conceived of the same mousetrap on December 31, 1990. Tweety never told anyone about the mousetrap and did not work on reducing the mousetrap to practice for many years due to financial reasons. Tweety finally actually reduced the mousetrap to practice on February 15, 2006. Because Tweety did not diligently work to reduce the invention to practice in the period before others' conception of the same invention, he is not entitled to a patent over Tom or Jerry.

However, if Tweety has published his idea before 2006, then this publication can be a basis to reject or invalidate Tom or Jerry's patent.
So I guess by simply mentioning his technique, PatimPatam has basically prevented anyone from filing a patent on the same thing, at least in the US. If there is already a patent about that it wouldn't be the case, but as I said previously it's all quite obvious and I really don't see how such a thing would stand up in a court.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Fredz wrote:So I guess by simply mentioning his technique, PatimPatam has basically prevented anyone from filing a patent on the same thing, at least in the US.
Looking at mahler’s post, i think it’s actually the opposite: i prevented anyone from filing a patent in Europe, but i still have one year to think if i want to try to patent in the US. Not that i would! (unless someone buys me off with stupids amount of money of course hahah).

Ok, thanx again for all the input guys, can we change subject now? :-)
User avatar
mahler
Sharp Eyed Eagle!
Posts: 401
Joined: Tue Aug 21, 2012 6:51 am

Re: POSITTRON: yet another proposal for positional head-trac

Post by mahler »

(IANAL)

And post some pictures from the 3d printed version of your Rift add-on so we'll having some more to talk about!
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Working on it :-)

I have been thinking about how to make the HW available for everyone, so it's both as cheap and as easy to mount as possible. The thing is you need a light but solid structure, and you need to be able to place the LEDs quite precisely in the right position; not sure if the foam core boards that MSat mentioned would allow that level of precision..

3D printing would certainly be ideal, but i'm a bit concerned about the cost. I think we can discard cheap materials like ceramics, polyamide, plaster or sandstone, so we would have to go with proper plastic.

I'm trying to minimize the volume (cm3) of the 3D model, while keeping it strong enough, in order to see if i can keep the cost reasonable, or if i should look for other options. For a 120 cm3 front section for instance i got a quote of 126$ on Sculpteo (http://3dprintingpricecheck.com), and you would still need to add the back section (about 60 or 80 cm3 probably) and the connection between both (not 3d printed), plus LEDs, etc.

Another option that has crossed my mind is usign a combination of "Lego Technic" and "Meccano" parts, but that ain't cheap either!

I'll post some pictures soon of what i have so far!
BillRoeske
Cross Eyed!
Posts: 102
Joined: Fri May 18, 2012 5:31 pm
Location: Houston, TX
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by BillRoeske »

If someone on here knows anybody that does plastic or fiberglass prop making (even hobby-level), they would be familiar with the process of creating a mold from a master that you could supply. From there, copies could be pulled relatively cheaply. At that point the tricky part would be finding people to produce said copies whenever there was enough interest and taking the bother to organize all of it.

So, overall, it would trade the cost and precision issues for distribution and management issues. :) Still, figured I'd toss the idea out there.
geekmaster
Petrif-Eyed
Posts: 2708
Joined: Sat Sep 01, 2012 10:47 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by geekmaster »

BillRoeske wrote:If someone on here knows anybody that does plastic or fiberglass prop making (even hobby-level), they would be familiar with the process of creating a mold from a master that you could supply. From there, copies could be pulled relatively cheaply. At that point the tricky part would be finding people to produce said copies whenever there was enough interest and taking the bother to organize all of it.

So, overall, it would trade the cost and precision issues for distribution and management issues. :) Still, figured I'd toss the idea out there.
You may get some good tips from the DIY "storm trooper" costume crowd (they like to use vacuum forming, which might be applicable here).

And here are some instructions for DIY mold making for duplicating fiberglass parts:
http://www.stevemeadedesigns.com/board/ ... iberglass/
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Thanks for the suggestions BillRoeske and geekmaster, and sorry about the late reply.. I think this could be a good option, especially vacuum forming looks like could be applicable here (more on that later in this post).


This last weekend I had a couple ideas as well regarding the HW building process and reducing the cost:

The first one is that i will probably need to change the initial design i made for the "rift-hugger", so that the main connecting piece between the front and the back sections is not directly supported only by the front part, but by the Oculus dev kit itself. This way the front section will support only it's own weight + LEDs (and maybe some batteries) and it won't have to be so strong and bulky. This will allow a potential 3d print to use less volume and therefore become cheaper, and at the same time open the possibility of using other techniques.


The other idea is that i will probably give different options for building the HW, so people can choose how much they want to spend / work on it.

The cheap option is probably going to be the same i used for my first prototype, simply strong cardboard! This will be possible thanks to the change on the design i mentioned. I will release the files with the templates, which people can print on cardboard paper and cut / glue themselves. This of course will be more work than 3d printing, but the main problem with this approach is that it won't be very precise. Because of this there will need to be some sort of calibration process after the main build, but it shouldn't be too complicated and you would need to do it only once.


The 2nd option will be 3d printing. I have been working on creating some proper STL files, that even if not final have allowed me to get some real quotes, and have a more accurate estimation of the final cost. Here is an example of the front section (17 x 10 x 4 cm bounding box, max thickness = 5 mm, min thickness = 2 mm):

Image

I have checked on a few 3d printing websites like sculpteo, shapeways or i.materialise.. here is the quote from shapeways for instance, using basic strong white plastic:

front section
104 cm3 - $88

back section
45 cm3 - $47

total $135, with international shipping included
($168 using black plastic instead)

Still a bit expensive for my liking, but it could be a good option for some people (especially if you have access to a 3d printer!). Also you could try the cheap method to start with and if you're happy with the results order a 3d print for long-time use (all the wiring and LEDs parts should be reusable without having to re-do them).


Finally i think the option mentioned by BillRoeske of making copies from a master mold is worth considering as well.. i personally don't have the time or equipment to get into this business, but if someone wants to print a 3d master and then make/sell copies using any method (like vacuum forming for example) that produces good results, similar enough to the original, i wouldn't be against it. Pure law of supply and demand! The only thing is that the LED mounting holes would probably need to be made "by hand" afterwards, so it may be necessary to run a calibration process after the assembly, in a similar way to the 1st method using cardboard.
German
Binocular Vision CONFIRMED!
Posts: 209
Joined: Tue Nov 23, 2010 5:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by German »

I should have my 3D printer back up and running in a few days. That model looks like something that can be printed via FFF(fused filament fabrication). I could give it a shot if you want, once you're ready.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

That would be great German, thanks. I would appreciate some feedback regarding general strength of the design and if there are regions that need to be reinforced for instance (of if on the contrary we can save material from some parts to make it cheaper, etc).

Having said that i must stress that this work is quite preliminary.. i won't be able to have a final design until i get the Oculus dev kit and i can take some proper measurements and run some experiments; also i'm not even sure if i'm going to use 5 or 8mm LED mounting holes for instance (still waiting to receive some more LEDs to test).

So basically you probably woudn't be able to really use it. If you're still interested anyway let me know and i'll PM you the STL file, no problem. If on the other hand you prefer to wait for something more final i'll understand!
German
Binocular Vision CONFIRMED!
Posts: 209
Joined: Tue Nov 23, 2010 5:18 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by German »

PatimPatam wrote:That would be great German, thanks. I would appreciate some feedback regarding general strength of the design and if there are regions that need to be reinforced for instance (of if on the contrary we can save material from some parts to make it cheaper, etc).
It looks fine, since it's just a truncated cube it should be pretty strong. It's hard to tell how the inside corners(where the truncations are) connect to the rest but if you used CSG, it should be fine.
Having said that i must stress that this work is quite preliminary.. i won't be able to have a final design until i get the Oculus dev kit and i can take some proper measurements and run some experiments; also i'm not even sure if i'm going to use 5 or 8mm LED mounting holes for instance (still waiting to receive some more LEDs to test).

So basically you probably woudn't be able to really use it. If you're still interested anyway let me know and i'll PM you the STL file, no problem. If on the other hand you prefer to wait for something more final i'll understand!
Well, when you want something physical and if you aren't happy with the cost of the commercial solutions, just let me know. I'd like to print a new case for my DIY Rift and it could be neat to integrate your solution right into it.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

German wrote:Well, when you want something physical and if you aren't happy with the cost of the commercial solutions, just let me know.
Thanks again, I will!
User avatar
mahler
Sharp Eyed Eagle!
Posts: 401
Joined: Tue Aug 21, 2012 6:51 am

Re: POSITTRON: yet another proposal for positional head-trac

Post by mahler »

Not sure if this has been mentioned before, but for people who would like to try this DIY design without the soldering, would a set of simple LEDs and batteries do the trick? http://www.metacafe.com/watch/502285/ho ... ky_lights/ - Apart from the batteries needing to be replaced pretty often, would these size LEDs even work?
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

mahler wrote:Not sure if this has been mentioned before, but for people who would like to try this DIY design without the soldering, would a set of simple LEDs and batteries do the trick? http://www.metacafe.com/watch/502285/ho ... ky_lights/ - Apart from the batteries needing to be replaced pretty often, would these size LEDs even work?
Pretty cool, yeah i guess it could work! The only problem i can see is that it would be a bit of a pain to turn on/off 16 LEDs each time you want to use it..
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: POSITTRON: yet another proposal for positional head-trac

Post by Fredz »

I guess it should be possible to connect them in parallel and use a 9V battery with some resistors to drive the whole. And add a single push button to switch them all on/off.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

Yep that was kind of the original idea i had, and what i'll probably end doing.. but it has the "problem" of having to connect the LEDs with resistors and wires (either by soldering or clamping for instance).

I believe mahler posted that video just to show that this method could be easier for some people, which is probably true; but obviously it has some disadvantatges as well.
geekmaster
Petrif-Eyed
Posts: 2708
Joined: Sat Sep 01, 2012 10:47 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by geekmaster »

Fredz wrote:I guess it should be possible to connect them in parallel and use a 9V battery with some resistors to drive the whole. And add a single push button to switch them all on/off.
Because multiple colors are used in this design, you cannot simply connect the LEDs in parallel. Different colors of LEDs have different voltage drops (typically 1.8v to 3.3v), and require different current-limiting resistors. You can also put LEDs into a series-parallel arrangement, so you do not need to waste battery power by dissipating excess voltage in current-limiting resistors.

There is a voltage drop (per color) chart on this page:
http://en.wikipedia.org/wiki/Light-emitting_diode

Ultra high output LEDs have even higher voltage drops (typically 2v to 5v).

For maximum battery life (increased LED efficiency), you might consider QLEDs (quantum dot LEDs), which have even smaller voltage drops and much less power consumption. QLEDs can be organically printed, making them a lot cheaper than silicon foundary production methods.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: POSITTRON: yet another proposal for positional head-trac

Post by PatimPatam »

geekmaster wrote:For maximum battery life (increased LED efficiency), you might consider QLEDs (quantum dot LEDs), which have even smaller voltage drops and much less power consumption. QLEDs can be organically printed, making them a lot cheaper than silicon foundary production methods.
Sounds very good indeed! But i thought this technology was still in development.. do you know of any online shop where you can already buy QLEDs? (had a quick look and couldn't find any)
geekmaster
Petrif-Eyed
Posts: 2708
Joined: Sat Sep 01, 2012 10:47 pm

Re: POSITTRON: yet another proposal for positional head-trac

Post by geekmaster »

PatimPatam wrote:
geekmaster wrote:For maximum battery life (increased LED efficiency), you might consider QLEDs (quantum dot LEDs), which have even smaller voltage drops and much less power consumption. QLEDs can be organically printed, making them a lot cheaper than silicon foundary production methods.
Sounds very good indeed! But i thought this technology was still in development.. do you know of any online shop where you can already buy QLEDs? (had a quick look and couldn't find any)
I only see products containing them, but no discrete QLEDs for sale yet. Not finding them in Google does not mean there are none. The right thing to do would probably be to contact the company that developed them.

You can read a little more about them here:
http://en.wikipedia.org/wiki/Quantum_do ... ing_diodes

This company is developing them:
http://www.qdvision.com/qled-technology

More about their use for LED lighting:
http://www.xconomy.com/boston/2010/03/3 ... -lighting/
A sign that QD Vision’s technology is gaining traction appeared this month when Charlotte, NC-based LED manufacturer Nexxus Lighting (NASDAQ: NEXS) announced initial production and shipment of its new replacement light bulbs, which use QD Vision’s quantum dot films. Nexxus says its so-called Array Quantum LED bulb fits directly into 400 million lighting fixtures already in place in the U.S. It’s the first time QD Vision’s quantum dots have turned up in a commercial product.
Other than that, most of the QLED research seems to be going into using them in video displays, due to their increased color gamut over LED or OLED displays, reduced energy consumption, and for their screen-printing viability.

It may be a bit early to use such new tech in something that only needs a handful of LEDs. Perhaps when ready for mass production it would be worth considering (for minimal power consumption). For now, just use readily available components.
Post Reply

Return to “VR/AR Research & Development”