Ann: Stereoscopic Sombrero Galaxy (2)

In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

On 2006-12-26, tontoko wrote:
In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense -- all the features
in an astronomical scene (except the spider spinning its web
in your telescope tube) are vastly beyond the focus range.
So using sharp vs. blurred features to reconstruct relative
distance won't mean anything. It might look pretty, but I wouldn't
say it was worth $50 for software to do that.

In other contexts this could be a cool idea, though.

Surely there isn't parallax enough to cause stereoscopic effect since
every star or galaxy has virtually "infinite" distance from the camera.


My software converts the dimness of the image to the distance from the
camera. Practically the galaxy or nebula is thought to have some
fractal structure and it causes blurry on the image taken by the camera

when the part of it is more distant from other parts.
The following image is an example of synthesized stereograph for a
fractal structure. As seen on it, more detailed, more distant it looks
like.

http://www.unmannedspaceflight.com/i...e=post&id=2049

Stuart Levy wrote:
On 2006-12-26, tontoko wrote:
In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense -- all the features
in an astronomical scene (except the spider spinning its web
in your telescope tube) are vastly beyond the focus range.
So using sharp vs. blurred features to reconstruct relative
distance won't mean anything. It might look pretty, but I wouldn't
say it was worth $50 for software to do that.

In other contexts this could be a cool idea, though.

On 29 Dec 2006 00:54:18 -0600, Stuart Levy wrote:

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense...

And if you look at the astronomical examples, that's what you see...
nonsense. The most obvious problem is that galaxies appear to be
floating in front of star fields, when in fact all the stars you see in
such shots are actually closer than the galaxies. But looking closely at
the galaxies themselves also reveals rather bizarre artifacts. There is
no connection between these images and reality. Some folks might like
the effect, but those same people might also enjoy viewing any image as
a pseudostereogram. You get a sort of 3D effect even when viewing two
identical images with the parallel or crossed eye technique.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

Stuart Levy wrote:
On 2006-12-26, tontoko wrote:
In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense -- all the features
in an astronomical scene (except the spider spinning its web
in your telescope tube) are vastly beyond the focus range.
So using sharp vs. blurred features to reconstruct relative
distance won't mean anything. It might look pretty, but I wouldn't
say it was worth $50 for software to do that.

In other contexts this could be a cool idea, though.

I agree. And it opens up the possibility that people will mistake this
for real data, the real thing, when it's just, at best, speculation as
to what the object would look like. We're already living with false
color, even in color photographs of astronomical objects done decades
ago, so why add to this mis-information?

And yes, I don't consider captioning a photo good enough, because most
people only remember the buzz words, and leave out those pesky details,
details which show the truth of the situation.

--- Dave

On Fri, 29 Dec 2006 17:36:28 GMT, David Nakamoto
wrote:

I agree. And it opens up the possibility that people will mistake this
for real data, the real thing, when it's just, at best, speculation as
to what the object would look like. We're already living with false
color, even in color photographs of astronomical objects done decades
ago, so why add to this mis-information?

I wouldn't compare this pseudo-3D technique with false color, and I
certainly wouldn't classify false color as "mis-information". False
color techniques enhance information content (even if some people
misinterpret the data). This 3D technique reduces information quality.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

I concede it is hard to confirm whether that stereograph is authentic
unless our left eye is 1000 light-year far away from right eye.

Chris L Peterson wrote:
On Fri, 29 Dec 2006 17:36:28 GMT, David Nakamoto
wrote:

I agree. And it opens up the possibility that people will mistake this
for real data, the real thing, when it's just, at best, speculation as
to what the object would look like. We're already living with false
color, even in color photographs of astronomical objects done decades
ago, so why add to this mis-information?

I wouldn't compare this pseudo-3D technique with false color, and I
certainly wouldn't classify false color as "mis-information". False
color techniques enhance information content (even if some people
misinterpret the data). This 3D technique reduces information quality.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

Please mind that my software can not reproduce the perspective with
100% authentication.
Firstly if the focal point resides between the foreground and
background, the software doesn't work properly because it can not
distinguish the object out of focus in the foreground from the
background.
Secondly (particulary in case of astronomical photos) if the object has
uniform appearance, the software can not estimate the relative distance
of it from other objects (therefore the
software does not give proper perspective for stars since every star
has similar appearance as a bright dot except sun.)

David Nakamoto wrote:
Stuart Levy wrote:
On 2006-12-26, tontoko wrote:
In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense -- all the features
in an astronomical scene (except the spider spinning its web
in your telescope tube) are vastly beyond the focus range.
So using sharp vs. blurred features to reconstruct relative
distance won't mean anything. It might look pretty, but I wouldn't
say it was worth $50 for software to do that.

In other contexts this could be a cool idea, though.

I agree. And it opens up the possibility that people will mistake this
for real data, the real thing, when it's just, at best, speculation as
to what the object would look like. We're already living with false
color, even in color photographs of astronomical objects done decades
ago, so why add to this mis-information?

And yes, I don't consider captioning a photo good enough, because most
people only remember the buzz words, and leave out those pesky details,
details which show the truth of the situation.

--- Dave

On 2006-12-29, tontoko wrote:
Surely there isn't parallax enough to cause stereoscopic effect since
every star or galaxy has virtually "infinite" distance from the camera.

Right.

My software converts the dimness of the image to the distance from the
camera. Practically the galaxy or nebula is thought to have some
fractal structure and it causes blurry on the image taken by the camera

when the part of it is more distant from other parts.
The following image is an example of synthesized stereograph for a
fractal structure. As seen on it, more detailed, more distant it looks
like.

http://www.unmannedspaceflight.com/i...e=post&id=2049

Hm. Even if you were using sort of fractal "texture" to determine
distance -- something that *could* be done independently of camera
focal length, etc. if the texture is actually there and actually
spatially uniform across an object -- consider the situation for
analyzing narrow-field deep-sky astronomical photographs.

Just from the fact that the things we're seeing cover tiny angles
on the sky, if they're roundish (as galaxies are), then the
ratio of near-edge-distance-to-us to far-edge-distance-to-us
must be nearly 1. Even for nearby Andromeda Galaxy,
maybe 5 degrees across at most, the near edge is only a few
percent nearer to us than the far edge. Most nearby galaxies
are only a few arc minutes across -- a few tenths of a percent
or less difference from near to far edge.

So, if you tried to estimate relative distance by comparing
the angular scale of some sort of texture (fractal or otherwise),
you couldn't get useful relative-distance results unless the
texture were very accurately measurable.

(In contrast, you *might* be able to get useful results if
you saw two galaxies of similar types, in the same field,
at very different distances. But you aren't showing any
examples like that that I can see.)

Stuart Levy wrote:
On 2006-12-26, tontoko wrote:
In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense -- all the features
in an astronomical scene (except the spider spinning its web
in your telescope tube) are vastly beyond the focus range.
So using sharp vs. blurred features to reconstruct relative
distance won't mean anything. It might look pretty, but I wouldn't
say it was worth $50 for software to do that.

In other contexts this could be a cool idea, though.

On 2006-12-30, tontoko wrote:
Please mind that my software can not reproduce the perspective with
100% authentication.

Well, I think the complaint is that it is reproducing the perspective
with 0% authentication. Its reconstruction of relative distance
is completely meaningless.

Firstly if the focal point resides between the foreground and
background, the software doesn't work properly because it can not
distinguish the object out of focus in the foreground from the
background.
Secondly (particulary in case of astronomical photos) if the object has
uniform appearance, the software can not estimate the relative distance
of it from other objects (therefore the
software does not give proper perspective for stars since every star
has similar appearance as a bright dot except sun.)

David Nakamoto wrote:
Stuart Levy wrote:
On 2006-12-26, tontoko wrote:
In the following website;

http://www.unmannedspaceflight.com/i...e=post&id=3392

the image shown is the stereograph of Sombrero Galaxy synthesized by
Stereographer (original image: HST).

For detail of Stereographer, visit;

http://139.134.5.123/tiddler2/stereo...tereograph.htm

Hmm. But, if the technique described here were applied to
astronomical images, it'd yield nonsense -- all the features
in an astronomical scene (except the spider spinning its web
in your telescope tube) are vastly beyond the focus range.
So using sharp vs. blurred features to reconstruct relative
distance won't mean anything. It might look pretty, but I wouldn't
say it was worth $50 for software to do that.

In other contexts this could be a cool idea, though.

I agree. And it opens up the possibility that people will mistake this
for real data, the real thing, when it's just, at best, speculation as
to what the object would look like. We're already living with false
color, even in color photographs of astronomical objects done decades
ago, so why add to this mis-information?

And yes, I don't consider captioning a photo good enough, because most
people only remember the buzz words, and leave out those pesky details,
details which show the truth of the situation.

--- Dave