Ballroom Dancing

I sometimes imagine a group of Facebook employees sitting around a table and asking themselves: "What exactly do a billion people do in VR?".  They are idly twirling an Oculus Rift headset on their polished conference table, looking back and forth between it and an open Facebook page, trying to plan their next move.

These are experts in human communication and behavior.  They have a pretty damn good handle on how their Facebook application is used and under what context.  Now what of VR?

I'm not Facebook, but if I were, I would be thinking about how to properly leverage time in VR.

Ok, lets run through the problem:

You build a beautiful VR ballroom and invite everyone to come.  Everyone does, they wander around the ballroom, dance a few waltzes and exclaim that "Why, this is simply splendid!  Isn't it dear?"

Evening gives way to night, night gives way to morning and eventually it comes to pass that everyone logs out.

As short time after, a college student in Osaka learns about the ballroom and logs in only to find everyone gone.  They look around briefly at the pretty environment and log out.

The next day, a past user in California wakes up and say to themselves, "That was a swell evening last night, I'd simply love to return once more."  

They log in, find themselves in an empty ballroom, sigh and promptly log out.

Log in, empty room, log out.  Repeat.

The dancing stops, the music dies, worlds crumble.  End scene.

--

Do you know what might sell social VR?

Let's call it "The Murmur".

The Murmur is the sound and motion of interaction between by people doing interesting things in the Metaverse.  

It is the stands in stark contrast to the deafening silence that permeates our traditional web.  It is the underlying reason why people eat at restaurants, drink at bars and watch movies together; and it may just be the key reason why people will be willing to tolerate the cumbersome trappings of modern VR.

VR will be attractive to the masses if it comes with the promise that the murmur will be waiting for them.

What if the ballroom never stopped?  What if the conversations and the dancers simply kept going.

Instead of depending entirely on live performances, VR worlds must transcend time and contain the essence and interactions of previous users.  The ballroom MUST always be filled with interesting people, even if they are long departed.  

Imagine if your Twitter timeline only showed tweets that were being posted now and nothing previous.  This is the current model for social VR.

Imagine if Bill Murray showed up in the ballroom for a couple of minutes and served a few drinks.  What if you could keep him on perpetually on a loop in the environment perpetually?  What if Bill could log back in later, say a few more quips and extend his loop?  

Just because you missed Bill last night, it doesn't mean you you can't see him the next day and just because he isn't logged in anymore, doesn't mean you can't try to interact with him.

You might think it would feel very wooden, but then you start to wonder what it might be like if Bill Murray's loop could also make eye contact with you at times and never quite looped the same way twice.  These aren't just recordings, these are self governing representations of self, designed to be consistently interesting and fresh.  

These aren't avatars as we know them today, but representations rendered with facial features and articulation that are very true to their host. 

 There's also good chance that Bill might also be appearing the shape of a walking narwhal or Kali.  Certain rules do not apply to where we are going.

"Don't think about your errors or failures; otherwise, you'll never do a thing." - Bill Murray

"Everyone loves Cinnebons." - Bill Murray

It's likely a bit confusing so let me walk you through an example:

• You log out for the night.
• A few minutes Bill logs in for 20 minutes, says a few clever things and serves a few drinks.
• The next day you log in and Bill, despite being long gone, is looped at the bar serving drinks.
• Being a big fan, you walk up to his avatar and tell him "Hi Bill, I'm a big fan."
• A few months later, Bill logs in and visits again, while he's there, at a given time, your avatar will walk up to him and deliver your previous interaction "Hi Bill, I'm a big fan."  
• If he responds, you'll see it happen the next time you log in.

The software layer, using cues from visitor interaction and opinion will automatically orchestrate a flow of recorded avatars and mix them in with the real visitors.  The longer other users gaze upon a recorded avatar, or stay within earshot, the more likely it will hang around and keep entertaining later visitors.  Literal "views" as a world governing metric.

A careful ballet of people and conversation needs to be orchestrated through software and the company that can pull it off with finesse stands to gain everything.

Easy to do?  Nope.  God no. 

But it will be an iterative process with lots of sample data to help designers along the way.

-- 

So.

If you missed your friends while you were logged out, no worries.  They walked up to your highly entertaining recording while you were gone and said a few choice things to you.  Now that you've logged in, their avatar will casually sidle up to you up and deliver those messages now.  

You can do the same for them.

(Note:  Expect a lot of advertisement delivered via this route as well, just thought I'd warn you now.  On the upside some of it will be fantastically entertaining.)

(Note 2:  A lot of what I've written above might sound unwieldy, but I strongly believe that we are moving into a time where all of our online personalities are intertwined with a guiding AI that helps smooth things along. Think how much smarter you seem to be with Google.  Why not enhance yourself in terms of being funny, clever and witty as well via a service?)

You might log in and have a number of avatars approach you over time, some will be real time users, many will be recordings of people you missed or people you don't know have on your friends list trying to reach you.  The best software services will be the ones that blend real time and past recordings into a logical mesh that "works".  

The murmur never stops.  The dance is without end.  You need never be alone again.

Give me a shout at @ID_R_McGregor on twitter, or send me an email at id.r.mcgregor@gmail.com.

Read More

Concave Dragons

The hollow face illusion in VR.

Michael Abrash has now given two very public presentations on VR that have focused on covering the brain's ability to process input and how misinterpretations happen along the way.  If you've not yet seen these talks, I strongly urge you to give them a look.

Time well spent.


Facebook's F8 Conference (March 2015):

Oculus Connect (September 2014):

One of the examples he recently gave as at F8 was a hollow face illusion  based off of a pretty well known paper dragon cutout.  If you take a look around the web, you'll find a number of different papercraft examples to let you print out and make your own.  Well worth the 20 minutes or so it takes.

Inspired, I wanted to see if this illusion translated into VR.  

Detailed  Scientific Answer:  Yep. 

Four takeaways:

1.  VR strongly supports the exact same visual illusions that trick us in the real world.  So, it is well worth studying these types of phenomena and seeing where they might be leveraged in VR.

2.  The effect is quite strong even in the absence of facial features (see last few moments of thevideo)

3.  VR makes for the perfect place to experiment with these kinds of visual tricks.  While working with paper models requires a bunch of fiddly scissor, tab, fold, refold, mountain fold, VALLEY FOLD, work you can duplicate and alter VR models in a few moments to try out different effects and theories.  Using this environment as a test bed, I imagine we are going to see a real a bunch of great new tricks and illusions whipped up over the next few years, discovered and refined in VR- then reproduced in the real world to devastating effect.

4. If you are working on a model, aligning edges and the model that you are working on deliberately plays with your ability to perceive concave / convex shapes, you will feel like you are losing your mind.

Edit:

I've had a number of requests for the source files for this project.  You can grab them here:  Unity 5 Scene.

Give me a shout at @ID_R_McGregor on twitter, or send me an email at id.r.mcgregor@gmail.com.

Read More

15 feet of fun

Let's look together at some things we might be able to do in a 15 x 15 foot space using Valve / HTC's Vive solution virtual reality gear.  This is pretty exciting stuff.

So here's what we've got to play with:

1 Room with dimensions of 15 x 15 feet.
2 lighthouse base stations (HTC / Valve)
1 headset  (HTC / Valve)2 handheld controllers (HTC / Valve)
1 test subject

<"I didn't agree to any of this." - Test Subject>

[Disclaimer:  "I don't actually have any of this equipment on hand, except for the room and I, ha ha, certainly am not keeping anyone in it and if I were, ha, ha, they'd be free to leave any time they wanted."]




------- Let's begin ----------------

If you've ever watched Little House on the Prairie, the show's credits roll with Laura happily running down a flower filled hill with their dog as the theme music plays.  Haven't seen it?  No matter, I've linked it here:

  


Now imagine that we wanted to experience this ourselves using the latest VR technology. We could actually get to be Laura running after Jack. Finally.

<"That's a pretty goddamn weird example you've chosen, I'm going home." - Test Subject >  

Right off the bat, there are a few problems.  The flower filled hill is very big and the path that Laura takes is longer than 15 feet.  Let's ballpark her target travel distance at say 150 feet.  

Target Simulation:

<"You screwed up your diagram, Jasper was the name of the Raccoon, not the dog." - Test Subject >   

In the real world, our test subject occupies a 15 x 15 foot room made of concrete. These dimensions are fixed and constant, we can't make holes in these walls despite our test subject's best efforts to do so.

VR Room + Target Simulation:

Starting at the edge of the room, if our VR test subject were to attempt Laura's run down the hill at speed, it would make it (at most) 15 feet before suddenly (and blindly) encountering the real world concrete wall.

<"Wait, why are you referring to me as 'it' "?  - Test Subject>   

 

Boy, we've covered a lot of important ground.  Let's sum up our scientific findings so far:

"If you try to run in a straight line for 150 feet from within a 15x15 foot room, you are going to run into a wall."

<"I'm actually quite badly hurt."  - Test Subject>

So, do we need to give up on the dream?

<"Yes, give it up, please, Jeez, I'm bleeding a lot."  - Test Subject>  

You can't give up on VR, I won't let you. Ever.

Well, let's simplify our goals and write down a list of wants.

Here are things that people want out of VR and VR Rooms.  This list is terribly important, so read the following with zest:




1.  People don't want to get sick when they move in VR.

2.  People quite unreasonably want to freely wander around for miles in virtual space while hardly going anywhere in the real world.


Huh. I only had two points, maybe this list isn't so important after all?  

Anyways, there have been some very clever and very determined attempts at solving these two "wants".  Here are some of them:

Just use your mouse, stupid.

The mouse was originally designed up as a tool to navigate business applications in the 1980s.  By all the accounts that I have read thus far, the folks at Xerox and then Apple were not trying to design a tool that was great at shooting people.  Yet, this is exactly what they did and what people started to do with it around 1992.

Pew, pew, pew.

Now, twenty three years later, the mouse is still, STILL the goto method for 3D world navigation.  Yes, yes, game pads have evolved considerably and now offer an absolutely amazing alternative*, but there's no denying that keyboard / mouse input does the heavy lifting. Would you doubt these guys?:

* no they don't.

This is wandering off topic, I know, I'm sorry, but it's absolutely baffling to me that we managed to arrive at such an ideal first person navigation tool while trying to make VisiCalc easier to use; and then what makes even less sense, is that we managed to do such a profoundly good job at it that we've not done much better in two decades of effort.

God knows we've tried to change the form factor and failed, again and again and again:

Too Gun.

Not Gun enough.

Too Microsoft.  (♥ you Microsoft, but Jesus, sometimes...)

I actually have a lot of grudging respect for this.

Unfortunately, using a mouse breaks the first "want" on our list (remember our important list?), our desire to not throw up while enjoying ourselves.  Yep, that's what we want.  

Wikipedia does a bang up job telling us why we have this problem with getting sick in VR:

In other words....... if your eyes tell you that you are moving (something that, wow, VR is really, explicitly designed to do...) but your body's movement/balance sensory system reports that you are actually NOT moving, or worse, moving in a different direction - some very old reptilian defense mechanisms kick in and conclude that you are likely hallucinating and have also likely been horribly poisoned.

  <"Wait, you said I did a bang up job explaining!  Are you... are you... actually... paraphrasing me?? Screw you!" - Wikipedia>

Your body then arms the torpedoes and makes ready to violently bring up anything you might have eaten and likely over your keyboard and your mouse that is so good at 2D, fantastic for 3D and absolutely crippling for VR.

Teleportation

Instead of walking around, let's teleport everywhere in VR space and dispense with all the problems involved with walking.  Let's just side step around the whole goddamn problem and enjoy ourselves for once.  

This is something that AltspaceVR pulls off rather well, take a peek at the video below if you'd like to see it in action.

This neatly solves the VR sickness issue since you are instantly jumping between points rather than smoothly translating between them.  This unfortunately does not even being to address our second "want", the desire to wander on our own two feet, so I'm going to simply have to disqualify on the grounds that it threatens to invalidate what I'm writing. NOTHING TO SEE HERE, MOVE ALONG!

Running in place

Virtuix Omni offers a platform (shown below) that allows a user use the sliding motion of their feet and directional facing of their body to navigate a VR world. 

This is a very bold product that tackles need desire to walk in VR head on.  I have tremendous respect for the audacity of launching such a solution.  From what I have read, it does a very good job at translating your movements into VR space and is a quality product.

Alas, I have also read that it does not entirely provide for our first "want", the need to not get horribly sick sometimes.  I haven't tried it yet to be sure, but some have reported issues. 

Via The Verge

I imagine that the full body motion does help to alleviate motion sickness for some, but we really do want to push for as close to 0% nausea as possible.  I don't think this is too much to ask from a fun experience, except for maybe the Gravatron.

Anyways...

Back to our 15 x 15 room.  In the basement...

<"Oh God no..... I'm still here....what day is it?".. indiscernible muttering - Test Subject>

Reportedly the HTC / Valve Vive Lighthouse tracking system gives us an effective way of accurately tracking your body in a 15 x 15 foot room.  So, far I've not come across any reports of the the device causing motion sickness. This is very promising!  

All changes to perception in the hmd are as a direct 1:1 result of your body moving in space. It might well be possible to wander around all day in VR space with one of these and not feel the least bit sick.  Fingers crossed.


Let's look about our small room and think about how we might escape these confines in VR.

<"Please, somebody, please send help. I don't know where I am or how I got here.  I can hear passing cars and... maybe birds... sparrows... was that a robin?" - Test Subject>

Let's run the Laura test again and see how far we get down the hill.

<"You can't be serious.  Oh God, you ARE serious. Didn't you write the results down the first time???  Nooooo!" - Test Subject> 

 

Not even close. Impact at 15 feet.  Yep, checks out with earlier findings.  Good to be sure.

Well.....um.. let's think hard now... what if.... instead of walking in a straight line, we had the test subject walk around the perimeter of the room after we throw some water on it to wake it back up?

Not only do we end up 90 feet short of our target distance, our test subject needs to make 3 90 degree turns in a simulation that is SUPPOSEDLY about running in a straight line.

<"That wasn't so bad." - Test Subject>

What if we had it walk in a circle instead?

<"I am not an "it" Goddamn it.  I'm a person just like you.  I have a family." - Test Subject>

<"At least I wasn't constantly making left turns, and the bleeding seems to have stopped." - Test Subject>

So despite our very thorough and very scientific investigation of BOTH squares and circles, both paths fall far short of our target distance and this is an obviously an unsolvable problem. I must conclude then, that VR is a bust and everyone is wasting their time.  

Thanks for reading!

<"What if I just, um, I could just .... walk around the circle again?" - Test Subject>

Wait, Hold on...  What if the test subject did multiple loops of the same path in order to cover the distance?  The pretty trees and flowers are an attempt to show a transition from one area to another.  You don't like my diagrams?

<"Your diagrams leave a lot to be desired, as do you." - Test Subject>

Ok... so as the test subject moves around the circle we update its perceived surroundings with new content.  So the trees give way to pretty flowers.  You can think of this kind of like a path winding up a mountain like a corkscrew, except there's no change in elevation, just scenery.  Maybe it isn't like a corkscrew.  I don't know.  

Here's a diagram anyways in case it is:

Note: Non-Euclidean Geometry could suddenly be a staple in the next few years.

<"Well done, you've created an infinite circular flower / tree path generator, great job! [slow clap]" - Test Subject>

Growing rapidly tired of circles and sarcasm, I've gone hunting for geometry that might help us out to start thinking about designing interesting levels.  Surprisingly my cultural ancestry came into play here.

If you start to look into "infinite paths", you'll soon come across Celtic designs such as the ones below.  These are knots of different kinds, lots of them bearing the name "eternity" or "trinity" or "endless".

So, you start to paw through these types of designs and start imagining them superimposed on the floor of your VR room as a paths to take that won't send you flying to walls.

<"Here. here." - Test Subject>

The best of them in my opinion contain gentle curves.

And we need quite a few more curves to make things interesting, and unfortunately a lot of the Celtic stuff gets a bit too dense and artsy.

 


The best stuff I've found, the good stuff, seems to dwell in what's known as "religious or sacred geometry".  I unfortunately don't have a background in it, but I can tell you about the symbolism but I CAN tell you that most of the designs are circular and a great many of them look like a good start when I think about using them as the basis of navigational guides for VR.

Here's a common motif that you'll find, albeit somewhat modified to my whims.  

Imagine this took up the floor of a 15 x 15 space.  Now imagine yourself walking along its lines and the path you might take.  Just hover your finger over the lines and consider the routes available to you.

You can travel this shape in a winding path, sometimes veering left, sometime veering right, sometime traveling straight only to branch off into another curve.  You could spend a long. long time navigating through this geometry while being gently steered clear of the walls.

<"I'm not getting up. I'm too tired. Find someone else.  I'm done." - Test Subject>

Level design for walk-able VR in a limited space is going to be challenging, kind of like authoring a geometric haiku.  

Here's my suggestion as we start to work on the problem:  




Turn off the scenery and embrace the void.



Even though it's a great film, I hesitate to invoke the Matrix here because you might die of cliche, but I do so only to mention that I think they actually gotten a fundamental right.  The sequences that take place in a formless white void ring true.  I believe we could find a shapeless void endlessly entertaining.... because um... we could put anything we want in it on a whim.

A void would allow us to explore some of the path finding trickery our geometric pathways offer.  We'd not need to worry quite yet about bending the world around us to fit the dimensions of our physical space as we simply remove all points of reference.  Besides, our processing power is better spent working on the few objects or people that have our immediate attention in VR.

You could wander in an endless sea of white, following faint geometric guides on the floor, or a floating way-point on the non-existent horizon.  Despite the lack of surroundings, you would always have a sense of where to go thanks to faint guiding paths on the floor.

Humans love paths, we do a really good job at following them, even at high speed on our daily commute to work. Lay down a path for a user in VR and you immediately give them something they intrinsically understand.  Watch young kids in a room with lines on the floor, they are magnetic and there is fun to be had in following them.

Give the user a path to follow, steer them away from the walls and provide interesting things along the way for them to find in the void.  Maybe we can start there.

<"Little help with this door?" - Test Subject>

Give me a shout at @ID_R_McGregor on twitter, or send me an email at id.r.mcgregor@gmail.com.

Read More