after watching the movie Avatar i found   on the internet ''cross eye'' 3D images.

in this link its explained everything. how you watch cross eye 3D pics.
www.neilcreek.com/blog/2008/02/28/how-to-see-3d-photos/

for 3D pics or 3D movies you need two images with different camera angle. like humans eyes see from a different angle. the difference is very small.

Inter Ocular Distance:
The distance between the centers of rotation of the eyeballs of an individual or between the oculars of optical instruments.

stereoscopic convergence:

The stereoscopic convergence is adjusted in stereoscopic imaging equipment to increase or decrease the stereoscopic convergence angle. Two imaging devices are disposed symmetrically relative to a median optical axis to generate video signals representing images of an observed scene. Depending on the required stereoscopic convergence angle, two image "read windows" are positioned symmetrically relative to the median optical axis, each being defined by a first pixel and a last pixel of a processed image line.

so if i understand this right now. inter ocular distance is the distance between our left and right eye. the bigger the inter ocular distance is the more 3D effect we have.
the convergence is the point in the 3D space where we want to point at. we decide from where on we have positive 3D effect and negative 3D effect?
am i correct?

the question now is how to use this in Poser 8?

do i make 2 cameras and then make them 2.5 inches apart? how do i control the convergence ?

Increase the 3D effect by increasing the yOrbit value.
Note - You may need to reverse the images right to left to get the proper effect.

cheers,
dr geep
;=]

geep its not so simple. for example in your render its very hard to watch the 3D. the guy with the white beard is closer to the camera. but to me it feels like the guy with black hair is forcing itself infront.

bagginsbill is a master in poser. i think he will find a way :) 

For the life of me I couldn't understand the camera measurements - trying to move it 2.5" using the camera options drove me mental ! Instead I just moved the camera to the right bit by bit then combined the images in PSP to see what result I got.

I'm not sure "orbiting" the camera along "Y" would work, as I understand it, you would be revolving the camera around a centerpoint. Eyes are next to each other, not set on a concave curve.

Anyway, I could be wrong, but here is the result of moving the camera to the right. Sorry, not sure how many units it was, I was just doing it as a test.

ETA: The second image is on the left.

Trivial.

Have the camera "point at" a dummy prop - any prop will do - I use a ball. This ball is the eye "focus point" - the position in space that the two "eyes" are looking.

You move this ball then hide it to keep the two cameras in sync.

Now, you can use your regular cameras, but I prefer to create a new pair of cameras for stereo. I make dolly cameras.

Position them both at 0,0,0 initially. Move one to X=k/2 where k is the inter-ocular distance you want. For exact human vision, use a realistic value. Move the other to X = -k/2.

Pick one to be the master, and parent the other one to it. This way, when you move the master camera, the slave will follow.

Have both cameras "point at" your focal target ball.

In Geeps render, he got the cameras reversed. Watch out for that when you put the images together.

For cross-eyed viewing, the image on the left is what the right eye sees.

Because he got them reversed, the central figure produces a visual paradox. From cues of a single image, he's behind the other two, but the stereo parallax places him in front. Swap them and all is well.

Quote - Trivial.

Have the camera "point at" a dummy prop - any prop will do - I use a ball. This ball is the eye "focus point" - the position in space that the two "eyes" are looking.

You move this ball then hide it to keep the two cameras in sync.

Now, you can use your regular cameras, but I prefer to create a new pair of cameras for stereo. I make dolly cameras.

Position them both at 0,0,0 initially. Move one to X=k/2 where k is the inter-ocular distance you want. For exact human vision, use a realistic value. Move the other to X = -k/2.

Pick one to be the master, and parent the other one to it. This way, when you move the master camera, the slave will follow.

Have both cameras "point at" your focal target ball.

did you just learn all of that in 5 minutes or have you already been rendering 3D before?

so if i understand right the convergence is the length from the camera to the dummy ball(point at). for example if i have a human and a car infront of him. if i point at the human then the car will be in positive 3D. if i point at the car then the car will be on the zero and the human in negative 3D? 

Just want to say thanks to bagginsbill !

Tried it like you said on some props as a test. So much easier that way, and the result was great.

Focal convergance never even occured to me, so thanks for that :)

Quote -
You can do this by rendering an image and then change the ...
camera's yOrbit parameter (very slightly) and then render a second image ...
... Like this.

Increase the 3D effect by increasing the yOrbit value.
Note - You may need to reverse the images right to left to get the proper effect.

cheers,
dr geep
;=]

Sorry, guess I got 'em bass ackwards, it was a 50/50 proposition and ...

Haste makes was ... er, makes 'em backwards.

It's sort of like ... reversed normals, no?! :lol:

cheers, (anyway)
dr geep
;=]

Epilog -
The yOrbit trick works when the center of the scene is in the center of the Universe.

Quote - Trivial.

Position them both at 0,0,0 initially. Move one to X=k/2 where k is the inter-ocular distance you want. For exact human vision, use a realistic value. Move the other to X = -k/2.

what does x= k/2 mean?

for example if k is 2.5 then i do 2,5/2  on the x dolly?

Quote - > Quote - Trivial.

Position them both at 0,0,0 initially. Move one to X=k/2 where k is the inter-ocular distance you want. For exact human vision, use a realistic value. Move the other to X = -k/2.

what does x= k/2 mean?

for example if k is 2.5 then i do 2,5/2  on the x dolly?

Dolly cam A 1.25 on x and Cam B -1.25 on x

I've done a few of these and they're always quite cool to see. This one has settings of x dolly 0.125, yrot -2 for the left image and xdol -0.125, yrot 2 for the right.

i noticed something. the camera that i move to the right needs to be used for the left eye? interesting isnt it?

If you move it to the right and then turn it to the left, it's functioning like the right eye -- which the brain reads as left. Go figure... LOL

i dont understand  :)

The brain takes the two signals from the eyes and reverses them -- TTTT, they go into the brain upside down and reversed, and the brain "fixes" it by putting the image back together. But for that to work, what's in the right eye gets "translated" into something that appears to come from the left.

When you think about everything involved, sight is a bloody miracle.

The cross eyed viewing means that your right eye observes the image on the left. In the stereo pair, put the image for the right eye on the left. It does not look like it is viewed from the left - it looks like it is viewed from the right. This is correct because the right eye is pointed at the left image when you look at it cross-eyed.

aha ok. so i render them out then switch them.

thanks. i will start posting renders. i noticed smaller renders are better becaue you dont need to force your eyes sso much.

I usually just convert these cross eyed pictures to analglyphs to see them easier.

i dont have the anaglyphs glasses.

Ice-boy, just cross your eyes until you see three images, look at the one in the middle...

Actually I prefer to view such stereoscopic images by relaxing the eyes rather than crossing them.. in other words, position the eyes as if you are looking right through the screen at something in the distance. I find it gives a more satisfying stereoscopic effect and is much more comfortable to hold.

This method of viewing means that the image for each eye needs to be on the same side, not opposite sides. I.e. image for the right eye is the one on the right.

It does mean that the images cannot, in practice, be further apart than your inter-occular distance, or you can't get the two images aligned correctly. Geep's images (both of them!) and the light bulb image work very well this way and produce excellent stereoscopic effects.

Geep - you simply reversed your whole image, which doesn't actually make any difference to the way it should be viewed. In order to view it cross-eyed, the left and right halves of the image need to be swapped over, but not flipped.

Also.. the giraffe is a pretty good example of the cross-eyed method. But the memorial is a bad example, because the two camera positions are too far apart, I think.

 This technique can be harmful to your eyes

Quote -  This technique can be harmful to your eyes

Which technique. crossing or relaxing?
And why?

cheers,
dr geep
;=]

Quote -  This technique can be harmful to your eyes

If you do it too much, sure. But given that we naturally cross our eyes to change focus from thjings far away to those near, I dont think it's worth worrying too much about.

thats why smaller renders are better. but you dont need to force your eyes so much. you dotn get the details but you get some nice 3D effects.

 Try this software - is java and has everything you need - inclusive the tutorials:

http://anabuilder.free.fr/welcomeEN.html

 4 years ago i made something stereoscopic for Poser. There is still a PDF here with some description: http://www.poserprofis.de

Good god, Dr. Geep is now making images fully 3D.  imagine if NaySay could pop out of the screen at you!  It's too horrible.

Can I disavow any knowledge of the above .......... puleeeeeeeease. :lol:

While poking through the bones of the now defunct poserpython.com, I happened across a python script to create a stereoscopic pair of cameras. The file downloads fine, but  I haven't yet tried it, myself.

web.archive.org/web/20061113101007/http://poserpython.com/dualcamerav1.zip

BTW, I get 50% more negative waves from The Nay Say Guy in when he's in 3D.  I need to go have a drink...

Whiskey works well ... :lol: ... That's what I use ... when dealing with da NSG.

seanmartin please excuse me for using your image.

Ahhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh

It's a NSG attack !

Everybody take cover, NOW !!!

My mind is going, Geep.  I can feel it... Daisy, Daisy, give me your answer do..........................

my first example.
the point of convergence is on the figure. for the inter ocula 2.5.

That looks great in "free-viewing" to me... "free-viewing" is what we used to call looking at stereo views by crossing our eyes. Effect might be better with the feet in shot. I found composition is an essential part of good stereo work back when I used to do Stereographic Photography.

i made now a test. i rendered the same image two times. the first one is the one that i posted. the point of convergence is on the figure. in the second render i put the point of convergence on the wall behind the figure.

i was reading somewhere that if you use differencet points of convergence that the 3D looks different. but in my examples it didnt look different. i forgot to save the renders so i can not post them.

does anyone ahe an answer?

Quote -  This technique can be harmful to your eyes

Quite the reverse in fact. It can be used as great exercise for the eye muscles. As said, nothing that the eyes don't do naturally anyway.

subtle stereo

Quote - Have you seen this effect? A jiggle print... works the way pigeons get 3D by bobby their heads.

seanmartin please excuse me for using your image.

No problem. That's pretty cool.

Quote - my first example.
the point of convergence is on the figure. for the inter ocula 2.5.

The problem with this, as I'm looking at it, is that the figure remains flat, like those late-generation Viewmaster reels.

Quote - Have you seen this effect? A jiggle print... works the way pigeons get 3D by bobby their heads.

Pigeons don`t see 3d by bobbing their heads, they see 3d by having two eyes.

• Pigeons don`t see 3d by bobbing their heads, they see 3d by having two eyes.*

Their eyes are on the sides of their head giving them a huge field of vision with very little if any overlap for parallax in the middle directly in front of them... bobbing the head gives them sequential parallax on those wide side fields of vision covered only by one eye. Or so I am told ;)  Sounds good to me in any case. An alternative explanation for why they bob their heads even when standing still is that they are just always listening to a really groovy soundtrack in their heads.

Pigeons bob to stabilize the visual field. As the body moves forward, the head holds still then quickly thrusts forward and re-establishes near zero motion as quickly as possible.

Pigeons don't bob their heads when standing still, nor when walking on a treadmill. They do bob when coming in for a landing in time with the wing beats, again to cancel their motion and stabilize the visual field on the retina.

http://www.loc.gov/rr/scitech/mysteries/pigeon.html

http://www.biomotionlab.ca/Text/TrojeFrost00.pdf