Celestial sphere image?

Hello, all!

As I've mentioned before, I'm making some animations to show Earth's
rotation, orbit, precession, and all that other fun rotational stuff.
Blame it on Oriel.

I've got everything done save for a star field background. Yes, I
know you can't really see the stars if the lighted portion of Earth is
also in-frame, but I want them anyway. Artistic license and all that.

Blender has a nifty utility that will generate a fairly good looking
star field background, and I'll use that if I have to. I'd rather
have an accurate celestial sphere, though.

For this to work, I need a rectangular image that I can coordinate
transform onto a sphere. In 3D modeling parlance, it's called UV
mapping--because you coordinate transform Cartesian x and y
coordinates into spherical coordinates u and v (with r assumed
constant). Think of it as a Mercator projection in reverse...
although it's not exatly the same thing.

Does anyone know of an image suitable for UV mapping onto a sphere
that would generate an accurate celestial sphere?

I suppose my only other realistic option would be to obtain a star
catalog and write a Mathematica routine to grab all stars below
magnitude m and project them on to a plane, then export the whole
shebang as a honkin' big jpeg or PNG that Blender can then wrap onto a
sphere.

Any other ideas on how I might approach this?

Thanks!
--
Dave

Dave Typinski wrote in
news
Does anyone know of an image suitable for UV mapping onto a sphere
that would generate an accurate celestial sphere?

I recently had the same need and I ran across the following:

http://www.oera.net/How2/TextureMaps.htm

The planetary textures are just fine, as long as you don't try
a low altitude fly by.

However, I found the star map to be of insufficient resolution
at only 2500x1250. This resulted in big blurry blotches of
stars when rendered at higher resolutions.

But then again, it may be good enough for your needs. Or perhaps
I'm just being nitpicky.

I'd love to find a similar UV map at a resolution at 4 times
that - 10000x5000. Or, figure out how to make one myself.

hmmm...this reminds me... I never did finish that animation
project. It had a similar purpose, to debunk a 'theory'.

Brian
--
http://www.skywise711.com - Lasers, Seismology, Astronomy, Skepticism
Seismic FAQ: http://www.skywise711.com/SeismicFAQ/SeismicFAQ.html
Quake "predictions": http://www.skywise711.com/quakes/EQDB/index.html
Sed quis custodiet ipsos Custodes?

Skywise wrote:

Dave Typinski wrote in
news
Does anyone know of an image suitable for UV mapping onto a sphere
that would generate an accurate celestial sphere?

I recently had the same need and I ran across the following:

http://www.oera.net/How2/TextureMaps.htm

That's great! Thank you!

I like his Earth Texture Atmospherical, too... the colors look like
they might actually work better than the one I got from this great
place, which seems to specialize in super-high-resolution stuff:
http://planetpixelemporium.com/earth.html

The planetary textures are just fine, as long as you don't try
a low altitude fly by.

However, I found the star map to be of insufficient resolution
at only 2500x1250. This resulted in big blurry blotches of
stars when rendered at higher resolutions.

I'm going to try it both ways: with this UV map and some fake ones I
made using Blender's "stars" background feature. The celestial sphere
would work fine for a view from L5, but it won't work for a view
looking down on the Earth from above the N pole. As the Earth orbits
in the model, the stars would seem to shift.

It's a model-space thing: Blender really doesn't like it when you try
to put something an infinite distance away, or even very large, for
that matter. My setup doesn't want to render anything further than
about 100 "Blender units" away from the camera.

....unless you parented the celestial sphere to the same empty that the
Earth was parented to, but didn't make it rotate. That would work!

Hm. I forsee more lilly-gilding to be done on this animation.

Yessir, there is definitely more than one way to skin a cat.

But then again, it may be good enough for your needs. Or perhaps
I'm just being nitpicky.

Not at all. It looks a bit dark, but some massaging with Photoshop
oughtta bring things up a bit. Or not. Great resource either way.

I'd love to find a similar UV map at a resolution at 4 times
that - 10000x5000. Or, figure out how to make one myself.

Well, if I ever get to it, I could write a Mathematica routine to do
it. All I need is a star catalog listing RA, Dec, Mag, and spectral
class. Still wouldn't grab any diffuse objects, though.

Come to think of it, with Blender's Python scripting ability, you
could probably parse the files and create the thing right in Blender.

hmmm...this reminds me... I never did finish that animation
project. It had a similar purpose, to debunk a 'theory'.

Oh? What theory was that?
--
Dave

On May 11, 1:57*am, Dave Typinski wrote:
Hello, all!

As I've mentioned before, I'm making some animations to show Earth's
rotation, orbit, precession, and all that other fun rotational stuff.
Blame it on Oriel. *


That's right, you equate a rotating celestial sphere with a stationary
Earth and its obverse, a rotating Earth to celestial sphere geometry
and this subhuman ideology surfaces in something you 'celebrate' and
consider as something actual -

1) There is no center of the universe, because expansion happens
uniformly everywhere.
2) There is a center of the universe, because it has a bounding volume

Six centuries after Archbishop Cusa recognized the need to discover
the motions of the Earth or rather,to find suitable arguments to get
rid of celestial sphere geometry,you and your colleagues can't get
enough of the astrological framework -

"And wherever anyone would be, he would believe himself to be at the
center.Therefore, merge these different imaginative pictures so that
the center is the zenith and vice versa. Thereupon you will see--
through the intellect..that the world and its motion and shape cannot
be apprehended. For [the world] will appear as a wheel in a wheel and
a sphere in a sphere-- having its center and circumference
nowhere. . . " Nicolas of Cusa

Any reader looking on would somehow know just how expensive this 'fun'
celestial sphere reasoning now is with no possibility of putting the
brakes on the reckless conclusion of carbon dioxide and global warming
by virtue of predictions but it something more fundamental than
that.How is it possible that in the 21st century that a person can
knowingly and without the expectation of any objection can discuss
celestial sphere things as actual instead of the horror that it is ?.I
am sure some know by now.







I've got everything done save for a star field background. *Yes, I
know you can't really see the stars if the lighted portion of Earth is
also in-frame, but I want them anyway. *Artistic license and all that.

Blender has a nifty utility that will generate a fairly good looking
star field background, and I'll use that if I have to. *I'd rather
have an accurate celestial sphere, though.

For this to work, I need a rectangular image that I can coordinate
transform onto a sphere. *In 3D modeling parlance, it's called UV
mapping--because you coordinate transform Cartesian x and y
coordinates into spherical coordinates u and v (with r assumed
constant). *Think of it as a Mercator projection in reverse...
although it's not exatly the same thing.

Does anyone know of an image suitable for UV mapping onto a sphere
that would generate an accurate celestial sphere?

I suppose my only other realistic option would be to obtain a star
catalog and write a Mathematica routine to grab all stars below
magnitude m and project them on to a plane, then export the whole
shebang as a honkin' big jpeg or PNG that Blender can then wrap onto a
sphere.

Any other ideas on how I might approach this?

Thanks!
--
Dave

Dave Typinski wrote in
:


I recently had the same need and I ran across the following:

http://www.oera.net/How2/TextureMaps.htm

That's great! Thank you!

I like his Earth Texture Atmospherical, too... the colors look like
they might actually work better than the one I got from this great
place, which seems to specialize in super-high-resolution stuff:
http://planetpixelemporium.com/earth.html

Thanks for that link. I'll check it out.


But then again, it may be good enough for your needs. Or perhaps
I'm just being nitpicky.

Not at all. It looks a bit dark, but some massaging with Photoshop
oughtta bring things up a bit. Or not. Great resource either way.

Now that you mention it, I also had to brighten up the image to
make it more useable. Had to be careful, though, or I started
seeing artifacts.


I'd love to find a similar UV map at a resolution at 4 times
that - 10000x5000. Or, figure out how to make one myself.

Just wanted to mention I just found the following at 8192x4096.
I found it using google's image search on "starmap" and 'extra
large' settings.

http://local.wasp.uwa.edu.au/~pbourk...ous/starfield/

The page references POVRay, which I happen to use. However, I
am using another program for the animation I'm doing. But hey,
UV mapping is the easy part, right?

This is a good one for our purposes as the stars at the top and
bottom edges are properly distorted for use in spherical mappings
of cylinders.


I also found this one, which is only 2880x1440 and unrealistic,
but really cool looking. It's also NOT properly distorted.

http://bb.nightskylive.net/asterisk/...hp?f=1&p=88466
http://www.geckzilla.com/apod/tycho_cyl_glow.png


Then I found this nice big 10000x5000 image. It happens to be
properly distorted, but it also has constellation lines. Some
time in an image editor could clear that up, but it's probably
not worth it since it's not much bigger than the first one above.

http://www.johnhpanos.com/starcal.htm
http://www.johnhpanos.com/skymap.jpg


hmmm...this reminds me... I never did finish that animation
project. It had a similar purpose, to debunk a 'theory'.

Oh? What theory was that?

Something to do with the Earth's axis tilting.

I'm just an amateur self taught CGI artist so I got kinda stuck
on a few things. But I have since acquired more tools so I should
be able to get around my problems now. If I can't render it right
the first time, I should be able to fix it in post production!

Brian
--
http://www.skywise711.com - Lasers, Seismology, Astronomy, Skepticism
Seismic FAQ: http://www.skywise711.com/SeismicFAQ/SeismicFAQ.html
Quake "predictions": http://www.skywise711.com/quakes/EQDB/index.html
Sed quis custodiet ipsos Custodes?

Skywise wrote:

Just wanted to mention I just found the following at 8192x4096.
I found it using google's image search on "starmap" and 'extra
large' settings.

http://local.wasp.uwa.edu.au/~pbourk...ous/starfield/

The page references POVRay, which I happen to use. However, I
am using another program for the animation I'm doing. But hey,
UV mapping is the easy part, right?

This is a good one for our purposes as the stars at the top and
bottom edges are properly distorted for use in spherical mappings
of cylinders.

Yes they are, and thanks a million!

I also found this one, which is only 2880x1440 and unrealistic,
but really cool looking. It's also NOT properly distorted.

http://bb.nightskylive.net/asterisk/...hp?f=1&p=88466
http://www.geckzilla.com/apod/tycho_cyl_glow.png

There's undoubtedly a way--PS plugin or standalone app--to transform a
"square" image like that into the warped/stretched format needed for
UV wrapping, although I haven't yet tried to find one.

Then I found this nice big 10000x5000 image. It happens to be
properly distorted, but it also has constellation lines. Some
time in an image editor could clear that up, but it's probably
not worth it since it's not much bigger than the first one above.

http://www.johnhpanos.com/starcal.htm
http://www.johnhpanos.com/skymap.jpg

Those lines aren't all that thick; it'd make for a nice
background--depending on the intent of the project, of course.

I emailed the author to find out more about how he got the
cyclindrical projection by stitching together a bunch of smaller
images. If and when that process becomes understandable, I'll post it
here.

hmmm...this reminds me... I never did finish that animation
project. It had a similar purpose, to debunk a 'theory'.

Oh? What theory was that?

Something to do with the Earth's axis tilting.

Seems we're doing the same thing. ;-)

I'm just an amateur self taught CGI artist so I got kinda stuck
on a few things. But I have since acquired more tools so I should
be able to get around my problems now. If I can't render it right
the first time, I should be able to fix it in post production!

Or just re-render it--which, unfortuantely, takes a while.

I'm not even a CGI artist, just self taught and stubborn enough to
wade through it all. I'm doing four animations, each one is 312
seconds long at 60 fps. I'll then use Vegas to make the final
composition. The down side is that I've 74,880 frames to render--and
each one takes about 35 seconds on this dual core Opteron box. That's
about a CPU-month of rendering out of Blender.
--
Dave

On May 11, 3:42*pm, Dave Typinski wrote:
Skywise wrote:

Something to do with the Earth's axis tilting.

Seems we're doing the same thing. *;-)


How intelligent does a person need to be to see that extending the
Earth's spherical geometry into the celestial arena via a calendar
driven Ra/Dec convention and without any qualifiers whatsoever will
produce a celestial sphere geometry ?.

Whatever drives people to believe that daily rotation to a stationary
celestial sphere does not retain the architecture of the apparent
motion about Polaris I do not know, but using a celestial sphere to
justify daily rotation has consequences which close to being a
dishonorable to human intelligence as it is possible to get -

http://www.youtube.com/watch?v=XTTDWhky9HY

In order to retain daily rotation to a rotating celestial sphere as
the reference for everything in the Universe and subsequently
rotational orientation to Polaris,they had to introduce a wandering
analemma Sun to blur the original distinction between daily rotation
and its true reference - natural noon.

Call it what you will,a nightmare or a crisis,the every-point-is-the-
valid-center basis for the so-called 'big bang' is simply Flamsteed's
horror come to life,don't take my word for it,try 600 years ago when
they tried to discover the exact arguments for planetary dynamics to
obviate the need for celestial sphere observations and conclusions
drawn directly from these observations -

"Suppose person A were on the earth somewhere below the north pole of
the heavens and person B were at the north pole of the heavens. In
that case, to A the pole would appear to be at the zenith, and A would
believe himself to be at the center; to B the earth would appear to be
at the zenith, and B would believe himself to be at the center. Thus,
A's zenith would be B's center, and B's zenith would be A's

And wherever anyone would be, he would believe himself to be at the
center.Therefore, merge these different imaginative pictures so that
the center is the zenith and vice versa. Thereupon you will see--
through the intellect..that the world and its motion and shape cannot
be apprehended. For [the world] will appear as a wheel in a wheel and
a sphere in a sphere-- having its center and circumference
nowhere. . . " Nicolas of Cusa

On May 11, 8:34*am, oriel36 wrote:

....don't take my word for it...

*****************

Feckwit, I don't think that will be much of a problem...

Dave Typinski wrote in
news
Something to do with the Earth's axis tilting.

Seems we're doing the same thing. ;-)

Well then. Seems I'll have ot finish up my work and then we can
compare notes.



I'm not even a CGI artist,

I should qualify that and say I use the term "artist" loosely.
A more accurate description might be CGI simulation.



just self taught and stubborn enough to
wade through it all. I'm doing four animations, each one is 312
seconds long at 60 fps. I'll then use Vegas to make the final
composition. The down side is that I've 74,880 frames to render--and
each one takes about 35 seconds on this dual core Opteron box. That's
about a CPU-month of rendering out of Blender.

At what resolution? My goal is HDTV 1920x1080, which take about 30
or so seconds per frame, but I do test runs at much lower resolutions
so it only takes me a few minutes to check a scene for results.

But, that's on a Core2 duo 3Ghz. I also have a quad and when I'm
ready for the final render I'll be doing a network render job.

If you want long rendering times, do a full raytrace with caustics
and index of refraction through complex glass objects. snooooooze

Brian
--
http://www.skywise711.com - Lasers, Seismology, Astronomy, Skepticism
Seismic FAQ: http://www.skywise711.com/SeismicFAQ/SeismicFAQ.html
Quake "predictions": http://www.skywise711.com/quakes/EQDB/index.html
Sed quis custodiet ipsos Custodes?

On May 10, 10:50*pm, oriel36 wrote:

How is it possible that in the 21st century that a person can
knowingly and without the expectation of any objection can discuss
celestial sphere things as actual instead of the horror that it is ?.I
am sure some know by now

*******************

And I am just as sure that you are just about the only person on the
planet who thinks this way.

It must be lonely at the top... very lonely...