Tilt-5 Was Magical | Hacker News Reader

Tilt-5 Was Magical

kguttag.com

153 pointsjpm_sd40 comments

Disclosure, I have done some consulting for Tilt-5.

I really don’t think people appreciate the technology of a retro reflective game surface and individual viewers. The concept that everyone gets to look at the game board from their own perspective is simultaneously “of course” and completely unintuitive to folks who are thinking “computer game.” I am not sure why this is but so often the comment from people who see the glasses for the first time is “Wow! I really didn’t know what to expect I guess, this is AMAZING!”

After it is out there, and people have them in their hands, I expect WoC to wake up and say “Holy crap, we could do MtG tournaments with this without renting a venue.” And I give Jeri a lot of credit for holding on to her vision of AR on game boards. That got out of her control at CastAR but here I think it is going to ship as envisioned and be marvelous.

The problem with hidden state is interacting with hidden state. Like if we're playing battleship the partition blocks you from seeing where I place my ships. But with a filtered display you can see where I move my hand and know there's a ship there. In short, hidden state requires hidden interaction which means some sort of controller.

At that point we might as well be playing video games on separate screens.

I think what we want is better board games. Which means cooperative play in a large world. Or maybe cool graphics. Or maybe the convenience of not having to buy/store physical games.

So excellent idea. Consider that the players are in different rooms/places potentially thousands of miles apart. You look down at a patch of ocean with your ships arrayed and floating as if at anchor. Using the wand, you move over the game board and a light grid it superimposed on the surface of the ocean. As you reach a grid you suspect might have a ship or boat you click the wand trigger. A projectile come in from off screen and either you see a splash in the ocean or an explosion. With some debris around it. You repeat with the grid square next to it and another explosion and then a flaming patrol boat visualizes and sinks before your eyes.

Meanwhile when your opponent moves, in your array of ships you see the ocean splash as they take shots that don't hit your ships, and when they do, your ship begins to burn and flame. Eventually they get enough hits that the ship will sink. At that point both you and you opponent get to watch it sink.

You cannot see your opponent or their hands so that state is always hidden.

How is that better than playing on a tablet?

The same reasons that playing in VR is better than on a monitor. But then one would question why this is better than VR. If all the players are physically separate and movement of physical pieces is limited (because all the opponent pieces are now virtual) you lose much of the benefit of AR.

This device is really best suited for people playing together in person (with online play a nice alternative, but not a primary target). As such I think a better example of the sort of hidden state it would work well with is Stratego. In that game the full board is visible to everyone but the identity of the individual playing pieces are not.

There are many other things you could do with hidden state, like show the stats of all your pieces floating above them, but not those of the opponents. Or show terrain of the board where your pieces have visited to you, but not to your opponents (who have to guess by your motions). It doesn't need to be completely hidden - partial knowledge or fog-of-war can make for interesting game dynamics as well.

If closely mimicking physical Battleship (or similar games that hide state by physical occlusion) is important, you could add a 2-retroreflective-sided board vertically in the middle to block each player’s view of the other’s hands.

But a software game has many other possible mechanics that don’t work with a physical plastic board and pegs, so more likely people will just modify the game so that seeing someone’s hand motions doesn’t reveal hidden state.

There is a controller. Unfortunately they haven’t yet revived the demo we did at CastAR where you reached out and manipulated real physical objects in the game. I have a video of that somewhere.

I think, AR like this is the future of gaming.

VR is a cool idea, but it falls short on so many levels. It looks cool to go into a space that isn't there, but you always have issues with the mismatch of your surroundings and the VR world.

Playing Xing, and the recent Myst remake is awesome on desktop, but very tiresome on VR. The haptics are obviously not there and either you can't really walk where you need to go or you "teleport" around which feels strange. Overall its tiring mentally and physically.

And the motion sickness, ugh. I'm not prone to motion sickness, but for some reason the subtle delays really affect me. I don't quite feel sick necessarily, but if I do VR for about 10 minutes, I feel just kinda off for about 3-5 hours afterwards. It's not a good feeling.

I honestly don't really think it will become a household thing until they develop black-mirror-type contact lenses.

I am somewhat prone to it, and I got seasick for about 12 hours from 30 minutes of sedate VR once. Eventually I realized what I needed to do was go outside and walk around to reset things. I haven't had anything that bad since- spending a bit more time in VR helps a lot- but it's not like I can just hand someone a headset and tell them to try it for more than a couple minutes without risking ruining their day.

This cracks me up. I don't think you are wrong, but I also remember not liking the original myst, thinking it was mainly a gimmick for cdrom drives.

So, not being wrong doesn't necessarily mean you are right. There is a chance the vr world lights up. Hard to really say.

As someone whose VR headset has been collecting dust for a year, I think the key is movement. Moving in VR just sucks. Either you teleport around which is disorienting or you slide with a stick which is nauseating. All my favorite VR experiences didn't require me to move further than my play area. If someone comes up with a cheap omnidirectional treadmill or something just as good it'll blow the market wide open. Until then, almost everything is more fun on a normal screen than in VR.

I found seated "sims" or games like Elite were fine for me, but anything that tried to replicate the FPS-style running around was a no-go. I guess it's even better in those arcade-style setups with moving seats to help replicate the feeling of movement and position seen with the eyes, but that's even more complicated than treadmills and such.

I think that for VR, the trick is to stick to experiences where your POV matches your meat-space position. Trying to shoehorn the conventions of flat-panel games into 3d VR either requires a lot of increasingly complex sim stuff or is just always gonna give you that mismatch between what you see and what you feel.

In that sense, I agree that AR is probably a lot more doable if you intend to be walking/running around.

So excited to have backed this myself. It's been...a long time coming. Not just from this kickstarter, but a past kickstarter.

The craziest part is that this is what it can do _before many people even have it in their hands_. The potential for this tech (outside of gaming even) is pretty far-reaching.

So glad that Jeri was able to get this thing to fruition this go-around amid the pandemic and all of the logistic problems that has created for creating new tech.

Could we put the reflector on a wall, blow it up, and use it as a window with virtual screens behind it to create a "multi-monitor" set-up like others are trying out in VR right now? Or do the tracking markers have a limitation in how big a rectangle they can follow?

> The resolution is a modest 1280 by 720p

> It has a 110-degree field of view

With that big of a field of view, and that low of a resolution, I don't think you'd be happy with reading text. But I'm just guessing.

That only applies to projection cone + pixels on the glasses. It's not how it necessarily translates to "effective" resolution on the wall when moving your head.

I mean, I don't typically read with the corner of my eye. The eye saccades are relatively central.

1280x720 AR eyeset can be good to excellent for reading text (I do it often).

But, the comfort is around a dozen lines (or less) per screen...

I suspect you are used to those 1280x720 being applied to a far, far smaller field of view. 110 degrees is remarkably wide.

Absolutely. The device is an Epson Moverio. I see through a quick search that the FoV should be 23°. And mind you, in some conditions the characters at the borders can lose focus or gain artefacts.

The display could sure gain from extra lines (a resolution of 1280x960, instead of 1280x720, could be optimal to my estimation) without causing visual issues, but potential visual issues with the left and right extremes are already evident.

I wanted to do that for CAD modelling if the glasses can be worn all day long. Look a the display for normal things, and look at the reflective panel for 3D things. I guess game developers could be interested too.

As others have said, that is probably not the best use of Tilt-5's technology. No matter how you move your head, it is still 1280x768 pixels spread over 110 degrees.

Doing a little math, that is only about 13 pixels per degree or 4.5 arcminute/pixel. Typical VR headsets are about 30 pixels/degree (2-arcminutes per pixel. For optical text reading like with a computer monitor, I consider 40 pixels/degree (1.5arcminutes/pixel) the bare minimum.

There is also some scintillation from the beaded screen. Not terrible, but enough that it would not be good for text. So even if the resolution was higher (which is possible with this technology), I'm still not sure it would be good as a computer monitor.

The case where it works best seems to be in "tabletop" applications, and it behaves best when the viewer is about 45 degrees to the surface. Business and military applications for things such as "sand tables" also fit this model.

Thank you for clarifying. And nice article! I had missed Tilt-5 and its predecessors completely so this was a very nice introduction

Do any of current VR headsets do reprojection within headset independently from the game framerate similar to what tilt-5 does? I would expect it to help a lot with minimizing any delays and thus reducing motion sickness. Shouldn't it work even better with VR since most objects are further away from eyes compared to virtual tabletop thus smaller change in shape of them when doing minor head movements?

I can think of two potential problems with VR reprojection that don't affect tilt-5 as much. Virtual tabletop rendered by most tilt-5 applications is closer to being flat surface which can be reprojected perfectly (ignoring pixel density). There is no rendered content to draw on the edges of screen when turning the head left or right, in case of tilt-5 unless your head is very close to the board nothing is rendered on the edges of FOV anyway since the board is in the middle of it. Might not be too much of problem since FOV edges need less details. Simply fading black or generating filler based on edge colors similar to TVs with ambilight might be good enough. There is also an option to render slightly wider FOV to give some buffer area if it means that hitting high framerate isn't as critical.

Yes, all of the headsets do some form of reprojection if frames are not generated fast enough, or to account for the latency between generation and display.

If you turn your head fast you'll see black at the edges where it's not caught up

Copied From michaelbuckbee [1]

Ellsworth is a personal hero of mine - incredibly smart, wildly talented and has a real vision for this space.

All that being said, it's a nightmare of a space which is why I don't think there's been a big funding event for Tilt5.

"Meta View" was an AR company that raised $75mil, had a star studded list of VR/AR technology folks, only ever shipped a couple thousand units and now is defunct.

Magic Leap raised $3.5 Billion and now has given up on shipping a consumer device (Enterprise only).

Microsoft's Hololens exited consumer applications even earlier, enterprise only.

Oculus Quest is the most successful consumer VR tech (about 5 million sold) but it's really unclear if they're anywhere close to turning a profit and they've spent tons to try and jump start game developers in VR.

Tilt5 would require from the ground up games to be made, large volumes of orders/units to be profitable and even if all that came together could still be kneecapped by chip shortages and supply chain issues.

[1] https://news.ycombinator.com/item?id=29322696

This would be a cool way to play a virtual version of M:tG or Hearthstone. Or Netrunner or Slay the Spire. Or Warhammer (Quest). Etc etc. Especially during a pandemic. But one can also already play RPGs where you have to use your imagination instead. Then you are the VR/AR.

Magic! That was the same feeling I had when I saw my first autostereoscopic display in about 1997 playing some DOOM-like game. I ACTUALLY LOOKED BEHIND THE DISPLAY LIKE I WAS A CAVEMAN TRYING TO FIGURE OUT HOW THE LITTLE PEOPLE GOT INSIDE THE BOX. O_O

I would love to work something like this into my D&D sessions. Looks amazing.

I didn't realise how precise the reflections from retro-reflective materials were. I'd always just thought they were for high vis gear, cool to see such a creative use for them here!

It's taken me a moment to think things through, but they only need to be precise enough to get back to the relevant user and not the one 90 degrees round the table.

The precision that's important is the projector. You project the top left 'pixel' to the top left of the board, it retro reflects back in your general direction from the top left of the board. Same for bottom right and all the other pixels.

I presume the better the retro-reflectivity the more efficient the system is (because less light is wasted going to places that aren't your eye), but it only needs to be 90 degree ish good enough for selectivity

There is a different image per eye, so much more precise than that

You don't need perfect separation to get a good stereo effect.

My son bought a pack of red-blue 3d glasses and we were looking at anaglyphic images online and enjoying them even though the images weren't perfectly clean.

In 3-d movies (Toy Story 3d) there are places where an isolated really bright object shows through cross channels and there some double imaging but it doesn't break the 3d.

These work for most people

https://en.wikipedia.org/wiki/Magic_Eye

and the channels aren't separated at all, your brain are doing it all out of desperation for closure. (Blood in the Gutter!)

Crosstalk can cause nausea and headaches in some people, just like VAC and motion lag.

That being said, circular polarization off of a retro reflective screen is one of the two state of the art ways of doing step stereo 3d: if you saw Avatar in a modern theater, there's a good chance that was used, so crosstalk should be minimal and much better than with an anaglyph.

[edit]

I should also note that circular polarization is different from the linear polarization that was used e.g. in Disney attractions in the '90s (and thus still used in some of the older attractions). With linear polarization if you tilt your head 45 degrees to one side or the other you will get an even mix of the L/R channels in each eye. Circular polarization does not have that issue.

Though it seems to me with the projector originating from the same plane as the glasses, circular polarization would be redundant. I'm sure the engineers there know more than someone who dropped out of physics after flunking the optics lab though :)

If you read the article, the retro-reflectivity is not used to separate left and right eye images. Two different projections with different polarisations are done, and then filtered into the correct eyes. See the section "Left and Right Eye Stereo Polarization" under "A little of the physics".

Wonder if you could "hack" it to work with retroreflective walls...

As an anecdote, the CEO of tilt five wrote the guidance software for the Astra rocket that went sideways when an engine quit upon takeoff.

That's how I remember the story, too, but I ran into Jeri at a conference a couple of weeks ago and she said she actually did the telemetry system. Which is also amazing.

oops, sorry for the misinformation

That was an amazing demo of Astra's guidance, in a positive way. Scott Manley did a video on it. Once the engine quit, the rest of the engines had to be vectored completely over to keep the rocket pointed up. And it did! If guidance hadn't been so adaptive, the rocket would have nosedived and blown up on the launch pad.

Reply on news.ycombinator.com