Joe Perez

iTrader: (8)

Build the ultimate stereoscopic head-mounted display. Counterbalance the CRTs with lead bricks, and mount it to a HANS device. Take it to Maker Faire and challenge people to wear it.

elesjuan

iTrader: (11)

Where the hell do you come up with these ******* brilliant ideas?!?!?!

Joe Perez

iTrader: (8)

What the hell else would you do with a pair of surplus 20" CRT monitors?

Actually, **** the lead bricks. Counterbalance the CRTs with a pair of Amiga 2000 computers, and challenge people to play one round of Dactyl Nightmare while wearing it. That'll separate the Men from the Miata Owners.



Sidebar: What ever happened to virtual reality, anyway? It was supposed to change the world. I mean, this was back in 1991, before some of you folks were born, and we were playing fully immersive videogames with 3D HMDs while standing in the middle of a spatial tracking rig! You crouch, and your character crouches. You look up, your character looks up. You turn right, and your character turns right. You move your arm, and your character moves its arm.

Call of Duty XVI is watered-down weaksauce by comparison.



Sidebar II: For those of you who clicked the link above, the Wikipedia article is wrong. Virtuality upgraded the CS1000 system to use a single A3000 per platform about halfway through the production run, but the original setup used three A2000HDs (stripped down to the bare motherboard) in the base of each platform. One per eye, and the third to run the game logic and interface to the server. I got to see them and chat with the technician when they had one of the platforms opened up for service at the installation in the Sarasota Square Mall.

y8s

iTrader: (8)

Virtual Reality pays my mortgage!

Not to steal away Rob's thread... but most of the virtual reality worth using is too expensive for consumers and is purchased by the military for training and simulation.

Believe it or not, DisneyQuest in Orlando used dual-CRT virtual reality setups for the Aladdin Magic Carpet Ride. Several years ago they decided it was time for an upgrade and we rebuilt them with LCoS based mini projection setups running at 1280x1024 resolution into which they pumped the original 640x480 video.

The new hotness now is SXGA OLEDs but they're not as bright as the LED illuminated LCoS so there are some limitations.

And to speak directly to Joe's concerns, we have a demo unit over at Sony.

Ford is also using VR for simulating car interior environment. it's all over youtube. they even showed our **** in one of their commercials once. I put a shitty copy of it on youtue.

Joe Perez

iTrader: (8)

That's exactly my point!

Today, VR exists only as something that we see on TV commercials, in military documentaries, or (gods help you) if you happen to own a copy of The Lawnmower Man on VHS.

But 20 years ago, VR was real. Any fool could walk in off the street, pay $2, and actually experience it for themselves.

What the hell happened? Technology is supposed to get cheaper and more easily accessible over time. My phone is a thousand times more powerful than the computers which powered the Virtuality platform. It can pinpoint its exact location anywhere on earth to an accuracy of just a few feet, and it can communicate with a hundred million computers scattered all around the planet at speeds that weren't even achievable over a distance of two meters by hard-wired connection twenty years ago. And yet the coolest game that I can play today involves flinging a spherical bird at a green pig with a slingshot?!

Something is wrong.

pusha

iTrader: (5)

what about portal? bitches love portal.

y8s

iTrader: (8)

go buy a shitty vuzix out of a skymall?

the problem with conventional VR is that it's as specific almost as a pair of glasses. aligning two displays is not easy for the average human and the error can make it less fun than 2D (and potentially require cleaning of vomit stains).

It's all about microdisplays. There's no volume in that market so there's little drive to improve or decrease costs. And you can't use phone-sized displays at 2 inch focal distance and achieve a result.

Every time we get some massive, deep pockets type in our office that wants to consumerize our stuff, we kindly shake their hand, show them a demo, and then explain to them why it'll never happen.

sidenote: you have no idea how unsatisfying it feels to be at a tradeshow and have a DARPA guy come up and say he wants virtual reality in a package like a pair of sunglasses in 3 years if you had "DARPA amounts of money" and you have to turn him away.

Joe Perez

iTrader: (8)

Whoa! You mean I can sit in a chair and push a plastic box around a flat surface with one hand and press keys on a keyboard with the other, while having to keep my head and eyes pointed straight ahead at all times so that I can look at a stationary rectangle sitting on the table in front of me?

I haven't heard of anything that revolutionary since King's Quest IV was released in 1988!



This is one point of confusion for me.

I understand that Nintendo had a lot of problems with user fatigue on the VirtualBoy, but I had thought that was more to do with the mechanical ergonomics (the unit had to sit on a table, rather than be worn on the head) and the fact that the displays were scanned-line-arrays (a single row of tiny LEDs and a set of spinning mirrors) rather than flat-panel displays.

By comparison, I don't recall any such problems with the Virtuality system. Granted, the HMD was a bit on the heavy side, and you also had to wear a backpack which seems curiously absent from all of the promotional photos, but this was an arcade game, designed to be operated by unskilled personnel and rapidly switched from one user to the next. I honestly don't recall the specifics of how the display was adjusted (it's been almost 20 years since I played one) but that's probably because it wasn't a major undertaking, just a single **** or slider or something to that effect.

Think back to the days of VHS camcorders, before flip-out LCD screens were commonplace. The viewfinders in those consisted of a tiny, high-resolution B&W CRT mounted perpendicular to the eye, with an angled mirror and a simple lens array. Adjustment usually consisted of a single thumbwheel. Professional cameras still use this technology. Now, I'm nearsighted as hell (-4.5), and I can easily pick up a camera, take off my glasses, and adjust the eyepiece in about two seconds. I can then hand the camera to a shooter with perfect eyesight and he can re-calibrate it for his eye in about two seconds as well. Ok, so with two viewfinders you'd also need an adjustment for pupil distance, but how big of a deal is that?

I understand that this is your business, and I don't question that there are clearly obstacles. My point is that what I'm describing has already been done, and it worked. All else being equal, any given technology is supposed to get cheaper, simpler, lighter, faster, and more easily accessible over time. Most do.

y8s

iTrader: (8)

I wish I had the chance to play with the virtuality stuff but I never ran across them up in the bay area. The closest we got was battletech with the little pods you sat in and played networked.



I suspect virtuality just wasn't making money on those setups. DisneyQuest has lots of spare units so they have minimal downtime. And those get beat up a lot.

OK so when I say "alignment" for the two eyes, it's not just interpupillary distance, it's yaw pitch and roll for each display. you can tolerate about 1/4 of a degree of rotational misalignment or vertical misalignment (it works out to 2 pixels on a 1280x1024 display with 60 degree field of view). you can tolerate about a 1/4 of a diopter accomodation for focus and convergence (if the focus is set to 5 meters, you wont get much eye strain if the convergence is about 2.2 meters to infinity).

Virtuality may have snuck by these issues by simply not letting any one person play for longer than a few minutes. And some people are more susceptible to the quality of the alignment. For the military applications or some Universities, they wear these things for 30 minutes or an hour and if the alignment is off, they'll have all sorts of nausea and headaches.

It gets even worse with partial overlap systems (middle FOV is stereo, periphery is not, but your field of view increases by 33%).

I just saw the specs on the Virtuality displays. 276x372 CRTs. Some of the later ones were 640x480.
this should be a fun browse for you:
http://www.cybermind.nl/Our%20Produc...20systems.html
oh wait, more rabbit hole!
http://www.cybermind.nl/Info/Visette_Pro_Tech_Specs.pdf

That's a pretty impressive field of view if you dont mind seeing giant pixels.

Our customers are shooting virtual Al Qaeda from helicopters or spotting planes while holding a rocket launcher. They need resolution.

By the way Joe... you are in the business of transmitting signals over the air. One thing we are ALWAYS after is a way to send low-latency video wirelessly. Can you get on that? ground soldiers dont like to be tied down.

Oh and if you're ever in the DC Metro area, we're finishing up our demo lounge so you can come play some video games.

elesjuan

iTrader: (11)

I still want to know where my flying ******* car is???? By the year 2000 we were supposed to have that ****...

Joe Perez

iTrader: (8)

I got to play the system at two locations, both in Florida. One was a dedicated installation in Sarasota, the other was inside an arcade in Gainesville. It was cool stuff.

So far as I can recall, the games were all multiplayer. Both of the installations I visited had two pods, but I've seen pictures of a four-pod setup. Always some combination of deathmatch / capture the flag, usually with an NPC enemy element as well. I may be wrong, but I think Virtuality probably invented the very idea of a multiplayer FPS - this was before Doom, before Wolfenstein 3D, even. If only they'd patented the concept, they might still be around today.

Cool stuff.

I expect you're right about the financial aspect vis-a-vis Virtuality. Those things had to cost a bloody fortunate to manufacture, another bloody fortunate to maintain, and they only built a couple hundred of them in total. And I'd expect that reliability / ruggedness might also have been an issue.

And frankly, the Amiga wasn't a good platform choice for that application. They were, without question, the best machines available for 2d bitmappped graphics (they had onboard bit blitters, which was unheard of at the time), but the Virtuality games were all polygon-based, and the Amiga was no better than a PC/AT at doing that. And, as a former Amiga owner myself, I will admit that the quality of the hardware, quite frankly, was pretty poor. Commodore cut a lot of corners on those boxes.


But I'm using the arcade machines only to illustrate a point. Obviously the arcade concept itself is more or less dead outside of certain special venues.

This technology belongs in the home.

Why not? All three of the major players in the home console market presently support some combination of "VR-like" position tracking using a combination of optical tracking and accelerometer-based sensing. They've already designed the input technology, so (and I ask this honestly) how hard would it be to apply that technology to a vuzix-like device? Microsoft has sold about a tetrazillion Kinects, Sony isn't far behind with the Eye / Move, and every single Wii comes standard with a motion-tracking controller. (The Wii design would probably be the best option for HMD use- put a couple of low-rez IR cameras on the user and sense the relative position of the base, rather than the other way around.)

Ok, so there's a liability concern with people walking into walls and whatnot. How many Wiimotes got tossed through flat-screens in the early days? They managed to get through that little problem. Just establish a virtual "cage" that the user must remain within, and pause the game when they step out of it.





Assuming that the HMD uses a pair of flat-panel display devices, and that both of those devices are rigidly mounted to a common surface, wouldn't that eliminate any possibility of axial misalignment or improper convergence?





How low is low? And what's the power budget?

For latency, it's unlikely that you will ever beat a composite analog modulator. There isn't enough bandwidth in the universe to transmit uncompressed digital video with anything resembling quality, so that inherently means coding latency. The higher-quality the codec, the higher the latency.

There's a growing hobbyist community centered around doing first-person-perspective telemetry for RC aircraft, and some of them are even using position-tracking HMDs and pan/tilt cameras. You might find some off-the-shelf hardware that's adaptable for your needs.

http://fpvpilot.com/default.aspx

http://eagletreesystems.com/osd/osd-pro.htm

http://www.hobby-lobby.com/rc_video_cameras_363_ctg.htm

http://www.hobbyking.com/hobbyking/s...idCategory=316



At the professional level, wireless camera links are starting to become common both for ENG and for sporting events, although they tend to focus on making the transmitter small and light while the receiver is big and heavy.

http://broadcastengineering.com/news...ngefp_cameras/


These guys also have a lot of cool stuff, and their tactical / law-enforcement products may actually be what you're looking for: http://www.vislinklaw.com/




Now that would be cool.

We actually just got done with the final phase of the WTOP build (they're in Georgetown), and that was the only project that we've done in that area recently. We did VOA, but that was about 12 years ago.

concealer404

iTrader: (3)

Did somebody say... "Virtual?"

y8s

iTrader: (8)

I keep pestering our EE to figure out how to extract stereoscopic information out of a single HDMI connector... I'll let you know.

You wish. Microdisplay manufacturers can't even maintain adequate tolerance on their panels. Remember, you're dealing with a few pixels worth of misalignment and the whole panel is a little bigger than your thumbnail. Tolerance in the panel, optics, and mechanical mounting add up. Pixels are like a few ten-thou per triad.



I confess, there are people who permanently mount the displays after aligning and just eschew the idea of adjustment. It's smart and maybe the way it should be. Good enough for 90% of people and Fair for 97%.

one frame at 60 Hz. Realistically, we could probably stand to have two frames. If you can't get below 50ms though, it's not worth a discussion.

Power for transmit is whatever the grid can handle. Receive is limited by battery power. As you know from your hipster trolling, your only limit there is weight. figure no more than a couple pounds of battery for 2 hours of time.

Bonus points for two way communication of tracking and comms.

WTOP. Traffic and Weather together on the 8s.

Joe Perez

iTrader: (8)

?

Full-rate stereoscopic video on a single cable became standard with the release of the HDMI 1.4 spec.

There are several different methods described in the specification. For gaming, the "frame-packing" method is recommended, where the left and right images are stacked top-to-bottom (at full resolution) in a double-height frame:



There are a couple of interleaved methods in the spec as well, which deliver left and right information sequentially, alternating either every other line, every other field, or every other frame.

You can download the 3d portion of the HDMI 1.4 spec for free by filling out this form: http://www.hdmi.org/manufacturer/specification.aspx (or you could check your PMs)




That's kind of what I mean. Align the device once at the time of manufacture, and then lock it in place with a few dabs of potting compound. I would imagine that the process could be automated fairly easily, using the same basic technique that's used to align BGA parts during the pick-n-place operation.




Hmm...

I honestly can't think of any products within the broadcast market that would satisfy the demand for 1-2 frame latency while operating at a sufficiently low datarate and being able to be received and recovered with a non-directional antenna. As a point of reference, the technique we use for OTA modulation has a transcoding latency of several seconds, and requires about 6Mhz worth of bandwidth. The systems used for wireless cameras at football games are a lot faster, but they typically require that a grip follow behind the camera operator holding a directional antenna on a stick and pointing it directly at the receive station at all times.

Check out the link I gave you to VisLink, though- they seem to be geared to exactly this sort of application.



Fun fact: WTOP is owned by Bonneville International, which is a wholly-owned subsidiary of the Mormon church.

y8s

iTrader: (8)

A directional antenna could probably fairly easily follow a tracked user with a motorized pan-tilt, could it not?

Check out this bargain:
http://www.brite-view.com/air_synchd.php

We're not HDMI, we're DVI over HDMI. But receiving and parsing a different resolution may not be impossible. The spec probably doesn't require HDCP or anything crazy, right?

Joe Perez

iTrader: (8)

You know, that's damned cool. I had no idea that wireless HDMI extenders existed. I was thinking more along the lines of industrial SDI. (We don't use HDMI at all within the broadcast industry.)

I'm honestly no HDMI guru, so apart from knowing that the basic formatting of DVI and single-link HDMI are essentially compatible, that's about it.

I do know that enforcement of HDCP is entirely up to the player. If the source device requests HDCP authentication, then the sink device must respond. If the source does not make an HDCP request, then the sink doesn't even have to know what HDCP is. Nothing about the HDMI spec requires that any device be HDCP-compliant; it's a completely separate set of standards, and is not even specific to HDMI. (HDCP can be implemented on DVI, DisplayPort, or even over an IP interface.)


So, when can I expect a stereoscopic motion-tracking HMD peripheral for my Xbox360?

TurboTim

iTrader: (9)

Why won't something like this work well enough until we get coolass HMD so we can actually turn around and such?

y8s

iTrader: (8)

You know about as much as I do about consumer video signals. Everything seems to carry DVI but that's the extent of my knowledge. I would think nobody cares about watching movies on a $16,000 HMD so HDCP is probably very irrelevant. Just upscale from 480p and deal with it.

We looked at SDI and while it was impressive, it was still too expensive to develop for wireless and never managed to reach our latency goals in the implementation we were shown. Plus it was huge.

Our take on consumer VR displays is (informally/unofficially) this: people wont pay more than they paid for a console for a peripheral that does more-or-less what a TV they already have does. And people will also not pay for a peripheral that excludes other people when the direction of gaming in the home has gone decidedly "Wii are a family".

But it could happen. There might be enough Joe Perezez with enough single-income-no-kids cash to spend a grand or two on something really nichey and awesome.

The nice thing about consumers is they are fairly tolerant of all the stuff the military and universities and industry is not.

Distortion is shitty? What's distortion?
Brightness is only 10 fL? What's a foot lambert?
800x600? that's better than progressive scan. and there's two!
50ms latency? my LCD input is 300!
and so on.

Reverant

iTrader: (10)

I find it disturbing that I remember that this guy's name is Lance Boyle.

Joe Perez

iTrader: (8)

I saw that video a year or so ago, and I'll admit that for something the guy rigged up at home it's extremely cool.

For a game which simulates a fairly stationary activity (being a helicopter gunner in Vietnam, for instance) that technology would probably work just fine. But I can't see it enhancing the experience of a game in which you control movement through the environment. All it's doing is substituting one form of input (the analog stick on the controller) for another. It does not add an immersive element.