The 3D depth was made to be about that of a typical nearby scene, almost
anyway.  You have to be kind of far away due to the image size depending on
screen res./monitor size if doing the diverging eyes method (on
3D_clouds_z.jpg).
Can be a bit painful for me unless 3 feet away.

Bob H.

"David Fontaine" <dav### [at] faricynet> wrote in message
news:3A3DA299.AAA1B452@faricy.net...
> It's not working very well for me...

Post a reply to this message

in povray.binaries.images, Bob H. says...
> I was trying out Kevin Ellis' clouds and so I figured I'd make a 3D stereo
> pair.
> I made the clouds with density_map using a zero density and regular density,
> also with both high and low cloud levels.
> Both diverged (z) and cross-eyed (x).
> 

Although I have been looking at quite a lot of stereo pairs in all sorts 
of ways with success, I have the suspicion that your pair does not have 
very much stereo information. What I mean is that it remains quite flat 
if I look at it (easiest is the x-eyed one for me). Also: when I subtract 
the one from the other using Photoshop, I get a nearly grey image either 
for the sphere or for the sky. Could you perhaps increase the effect a 
little to see if that makes it better for me? The scene should be very 3D 
in my opinion!

-- 
Regards,  Sander

Post a reply to this message

"Bob H." wrote:

> The 3D depth was made to be about that of a typical nearby scene, almost
> anyway.  You have to be kind of far away due to the image size depending on
> screen res./monitor size if doing the diverging eyes method (on
> 3D_clouds_z.jpg).
> Can be a bit painful for me unless 3 feet away.

Tried it, better, but still kinda flat.

--
David Fontaine  <dav### [at] faricynet>  ICQ 55354965
My raytracing gallery:  http://davidf.faricy.net/

Post a reply to this message

Not too much trouble at all to redo.

Any better?  The camera separations are only 5 units different from last time
but it did look better to me.  Any more and I think it wouldn't have the right
depth to it.
Remember the sphere doesn't shift, it's moved along with the camera.  Again,
x3D is the cross-eyed and z3D is the diverged pair.

Bob H.

Post a reply to this message

Attachments:
Download 'z3D_clouds.jpg' (38 KB) Download 'x3D_clouds.jpg' (38 KB)

Preview of image 'z3D_clouds.jpg'

Preview of image 'x3D_clouds.jpg'

The clouds look right, but they still appear to be raised out of the sphere.
I'd say this is because you're moving the sphere with the camera, which
distorts the effect.  You need to keep the sphere in the same position, and
the camera look_at position the same, but move only the camera to get the
right effect.

--
Lance.

http://come.to/the.zone

Post a reply to this message

in povray.binaries.images, Bob H. says...
> Not too much trouble at all to redo.
> 
> Any better?  The camera separations are only 5 units different from last time
> but it did look better to me.  Any more and I think it wouldn't have the right
> depth to it.
> Remember the sphere doesn't shift, it's moved along with the camera.  Again,
> x3D is the cross-eyed and z3D is the diverged pair.

Thank you for making the scene all for me! It is slightly better, and 
perhaps my eyes are starting to fail me just a little (or the brain - 
that's where the actual merging of stereo pairs occurs, I think). I still 
have the idea that there is something about this scene, resulting in less 
3D effect than one would expect :)) Now don't get mad at me :-))
-- 
Regards,  Sander

Post a reply to this message

in povray.binaries.images, "Lance Birch" <-> says...
> The clouds look right, but they still appear to be raised out of the sphere.
> I'd say this is because you're moving the sphere with the camera, which
> distorts the effect.  You need to keep the sphere in the same position, and
> the camera look_at position the same, but move only the camera to get the
> right effect.

This could just be true. If you move the sphere with the camera, and the 
clouds are far away with respect to the distance camera-sphere, the 
reflected clouds could seem to lie flat on the sphere's surface. (?)
And: there wouldn't be much 3D in the clouds themselves, I think.

-- 
Regards,  Sander

Post a reply to this message

Yes that's right, you see by moving the sphere, it alters dramatically the
reflection that the other eye would normally see in the sphere.  The
reflection is actually so different that you see the sphere as having the
reflected clouds sitting "out" of the surface.

Instead you should move the camera like this... (warning, bad ascii art
ahead)

    o
  /   \
 /     \
l      r

Where the o is the sphere (or the camera's look_at point which should be
centred to the sphere, or top of sphere) and the l and r are the left and
right cameras.

Nothing should move in a stereo scene except the camera, or you won't get
the right effect.

--
Lance.

http://come.to/the.zone

Post a reply to this message

Uh, guys... I intentionally moved the sphere with the camera for the
possibility of it being an odd affect to begin with  :-)
But trying with a look_at that isn't parallel sounds like a thing to do too.
Were this done in the usual fashion, as you outlined there Lance, then the
sphere needs to be huge and the camera very distant from it, also requiring the
clouds to be scaled upward.
I'm not so sure real clouds have much of any parallax worth making a stereogram
of though.  And that's what would happen if I left the sphere/camera distance a
typical ball-size, a much smaller left to right shift would need to be used.

Bob H.

"Lance Birch" <-> wrote in message news:3a416509@news.povray.org...
>
> Instead you should move the camera like this...

Post a reply to this message

Welcome to a changed render anytime, within reason  :-)
If you see or saw my reply to Lance B.'s message then maybe it will make more
sense.

Bob H.

"Sander" <san### [at] stolscom> wrote in message
news:MPG.14aaf7b5e7718b909896c5@news.povray.org...
>
> Thank you for making the scene all for me! It is slightly better ---

Post a reply to this message