Interesting Idea From XKCD

[Devonai]

This XKCD comic suggest a rig for 3D cloud viewing, and also mentions the problem of depth perception for astronomical observation.

http://xkcd.com/941/

I've been thinking about the possibility of doing this for astronomy, and I don't think it would work.  I think the parallax would be too small for any perceivable results.  I was wondering which object would be best for camera convergence but I don't think it would work for anything but the moon.

Any thoughts?

[MicroBalrog]

How hard would it be to program the two cameras to create a 3D image like the one created by the twin cameras of the LG 3D phone? And then, uh, feed it into the screen of the LG 3D phone?

[birdman]

Wouldn't crossing the cameras over like shown yield an odd result?
(in the diagram, the right camera is in front of the left eye, and visa versa)

As for astronomy, well, the unit parsec (a little over three light years) is "parallax second of arc" meaning an object viewed from opposite sides of earths orbit around the sun (roughly 186 million miles) will differ in apparent position by one arc second.  Now, a human eye's resolution is about 60 arc seconds, so that means with a 60-80x telescope, viewing images six months apart, you just might be able to determine the parallax of the nearest star.   Basically...moon, yes, but stars are far.

One arc second (our 60-80x telescope) of parallax on the moon would require a baseline (separation between cameras) of about 3/4 of a mile.  So if you put two web-cam equipped spotting scopes that far apart, (or farther, you would need 100-150 miles to see the moon as a sphere) you could see it in 3-d...barely.  Anything else is too far, and why we stopped using parallax to judge celestial distances beyond a few 10's of light years.

[Devonai]

That's what I figured.  I guess I'll stick to using Celestia to zoom around.

[CNYCacher]

I've known about separating the cameras in stereoscopic photography to increase the depth perception effect, but never heard of it done to this scale.  Pretty cool

[AJ Dual]

http://dorigo.wordpress.com/2006/02/16/libido-at-40/

Expensive, but dual-mounted refractors made into an overgrown set of binoculars will actually give you some of the "feeling".

[Regolith]

http://dorigo.wordpress.com/2006/02/16/libido-at-40/

Expensive, but dual-mounted refractors made into an overgrown set of binoculars will actually give you some of the "feeling".

That's not a set of binoculars.

Now THAT'S a set of binoculars.

[freakazoid]

What could be so interesting on the floor? 

[Regolith]

What could be so interesting on the floor? 

They're reverse binoculars.  They're so big, that you have to look down the opposite way the binoculars are pointing, because it's too much effort to get them up high enough to look through them the normal way.

[AJ Dual]

Well, they're a pair of Dobsonian reflectors, and you always have to stare into the eyepiece near the top aperture of those.

Cool concept, however, extraordinarily limited potential to find and track objects if you have to drag the chair around and stand up/squat etc. depending on the elevation of the object you're trying to view.

[Regolith]

Well, they're a pair of Dobsonian reflectors, and you always have to stare into the eyepiece near the top aperture of those.

Cool concept, however, extraordinarily limited potential to find and track objects if you have to drag the chair around and stand up/squat etc. depending on the elevation of the object you're trying to view.

That, too, but they could have just had the eyepieces pointing the other way (same position, just have the 90 degree angles rotated 180 degrees the other way).  Except for having to make sure everyone's head would fit between the telescope bodies, and the fact that they'd have to mount it up higher, which would take more effort to set up and would be much less stable.

Or they could have gone with a Nasmyth telescope, which would eliminate mounting problems that doing that with a Newtonian would.  Only problem would, again, be the unstable high mount and a longer focal length, the latter of which isn't necessarily a bad thing if you aren't taking pictures through it.

[AJ Dual]

I've seen some really impressive binocular rigs where the chair for the observer is part of the whole T&E mount.

[Scout26]

At least the neighbors won't think you're looking in their windows.....

[230RN]

First time I got a sense of depth perception in space was when I watched Jupiter's moons orbit over several periods.  I really got a sense of the little balls swinging around on strings.

[CNYCacher]

Putting the cameras 100 yards across simulates the scale of a person with a face 100 yards between eyeballs.  That's enough to make the clouds seem smaller.

How far apart do the cameras need to be to allow you to perceive depth between the stars?

Lets take as a round number 3 inches as a normal eye separation.  If we put a camera on either side of the country, roughly 3000 miles apart, the scale would seem to suggest the moon could appear to be (200,000 / 1000) = 200 inches away, and the rest of the stars would seem to be out in infinity I imagine.  At that scale, mars could appear as close as (36,000,000 / 1000) inches, or just 3000 feet way!  That is certainly close enough for your brain to perceive depth against the stars.

Let's see, what is the moon in diameter?  Roughly 2200 miles?  If we follow our 1000 miles to an inch scale, the moon would appear the size of a say, a tennis ball hovering 15 feet away.  Mars could be a street light as close as under a mile, and the rest of the stars would probably still be hung on their blanket out in infinity.


Here's another idea, take two images of the stars 6 months apart.  The distance between your eyes would be roughly 183 million miles.

Nope, nope, that's not right.  Different sets of stars would be visible at those times.  You would have to go it at the spring solstice and the fall solstice, one at dawn and one at sunset.  I'm not sure you can get stars to  appear on camera though at those times.  Perhaps if you were up high enough in elevation?

Well, I guess you could go like 3 or 4 months apart instead of six, that way you can time your shots for nighttime.  1 month after fall solstice and shoot at 8pm local standard time.  Then come back 1 month before spring solstice and shoot at 4am local standard time.  I figure you' could just shoot straight at Polaris for both shots, rotate one of them 180 degrees.

Hell I am doing this!