It is currently Sun, 17-12-17, 8:14 GMT

All times are UTC




Post new topic Reply to topic  [ 16 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: Mon, 23-01-12, 20:30 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
Hi all,

The idea of using a hires panoramic image of the entire MilkyWay as a realistic rendering background in Celestia, was pioneered by Guillermo Abramson recently:

http://www.shatters.net/forum/viewtopic ... 23&t=16697

His nice add-on is available at

http://www.celestiamotherlode.net/creat ... lkyWay.zip

His texture is based on a panoramic view of the whole sky credited to
ESO/Serge Brunier, http://www.eso.org/public/images/eso0932a/.

+++++++++++++++++++++++++++++++++++++++++++++++++
Unfortunately, there is a major, persistent problem with alignment!
+++++++++++++++++++++++++++++++++++++++++++++++++

After spending quite a lot of time with it, I was able to prove that the image from the ESO site is badly distorted (internally), making a decent alignment within Celestia(.Sci) basically impossible (no matter what projection and/or coordinate range is assumed). So I contacted Serge Brunier, who confirmed these inherent distortions by email. Possibly, the reason was that he assembled his 224 individual frames with software, which --according to Serge B.-- does not have astronomical precision standards.

Since the inherent misalignments are really disturbing, I started working instead on Nick Risinger's panoramic MilkyWay image, http://skysurvey.org/, that is also publicly available under a Creative Commons License up to 3000 x 1500 pixels resolution.

Here is a small window from the 3000x1500 original, so you can tune yourself in on the amazing amount of details and subtle colors:

Image

Despite its smaller resolution, it has a number of major advantages, notably:
  • I achieved a close-to-perfect (!!) alignment out to +- 35 degrees north/south of the galactic equator, such that even the further-away nebulosities fit perfectly with the Celestia(.Sci) stars/globulars (Pleiades, Orion, Antares,M4, ...)! Furthermore, I replaced Guillermo A.'s (very) approximate sphere mesh with only 384 vertices by a considerably smoother one by Selden Ball with 703 vertices,

    http://lepp.cornell.edu/~seb/celestia/mw-msx.cmod.

    Note: Unlike Guillermo's MW texture, neither a shift of the central meridian nor a vertical flip of Nick Risinger's image is required for the alignment! Hence, don't mix parts of the two add-ons, without knowing precisely what you are doing!!

  • Due to a way more expensive photographic toolset,
    Serge Brunier wrote:
    I didn't use a 100 000 USD material as Nick to do my picture, I only used a Nikon Camera...

    Serge

    Nick Risinger's MW image exhibits most well-known red H-alpha glows all over the picture, as well as other nebulosities (Orion, Antares/M4,...). These provide great views in conjunction with Celestia(.Sci). Let's have a look at some examples:

    H-alpha glows and other nebulosities:
    ============================
    The shown Orion field is filled with red glowing hydrogen gas: in the north, the spherical nebula Sh 2-264 around the star lambda Orionis, in the lower left the huge arc called Barnards Loop, in the lower center at the three belt stars the well known IC 434 (including the prominent horsehead dark nebulae), and below Orions sword region with the great Orion nebula M 42/43.

    Image

    or with guides
    Image

    Next, let's look at the Antares/M4 region:
    Note that the actual visibility of these nebulous regions depends on your monitor adjustments!

    a) close-up:
    Image

    b) at smaller zoom, showing the dark lanes etc...
    Image

  • I put all my image manipulation know-how in to produce a star-cleaned version of Nick's image that I think can hardly be improved given his largest available resolution of only 3000x1500. I am still waiting for an answer from Nick, about making a free 4096x2048 size available for Celestia(.Sci) under a CC license. Of course, much higher resolutions do exist, but they are not to be circulated freely in the net, since they form the basis for a solid wallpaper selling business.


Last not least, here is an overview of the MilkyWay in Celestia(.Sci). Since Nick Risinger's original involves rather strong colors and high brightness, I have reduced both parameters by 50% which appears like a good choice.

[click on image by all means!]

Image

Finally: Here is my add-on for download

+++++++++++++++++++++++++++++++++++++++++++++

http://www.celestialmatters.org/users/t ... singer.zip

For details, license, installation,... see README.txt! Alignment data are in mw.dsc.
+++++++++++++++++++++++++++++++++++++++++++++


Enjoy,
Fridger


Last edited by t00fri on Sat, 28-01-12, 21:05 GMT, edited 4 times in total.

Top
 Profile  
 
 Post subject:
PostPosted: Mon, 23-01-12, 21:47 GMT 
Offline
User avatar

Joined: Mon, 07-01-08, 13:30 GMT
Posts: 339
Location: Switzerland
Very nice this one, thanks for providing it! I'd say it's worth putting this on the front page... last thing to go there was my W UMa add-on back in 2009 (!!!)


Top
 Profile  
 
 Post subject:
PostPosted: Mon, 23-01-12, 22:04 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
ajtribick wrote:
Very nice this one, thanks for providing it! I'd say it's worth putting this on the front page... last thing to go there was my W UMa add-on back in 2009 (!!!)


Thanks for the flowers...My God, I first have to remember how to make a frontpage for the CM web site ;-) My last one was done by Runar, ahem. I'll think about it, perhaps it's not a bad idea.

Fridger


Top
 Profile  
 
PostPosted: Tue, 24-01-12, 14:19 GMT 
Offline

Joined: Tue, 24-01-12, 12:24 GMT
Posts: 3
Thanks Fridger, for the excellent update of Abramson's delightful addon. The creation of realistic views is now much enhanced.

Many thanks for all your work most of which, unfortunately, goes over my head.
David.
PS I've just joined CM.


Top
 Profile  
 
PostPosted: Tue, 24-01-12, 14:38 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
DoJoMo wrote:
Thanks Fridger, for the excellent update of Abramson's delightful addon. The creation of realistic views is now much enhanced.

Many thanks for all your work most of which, unfortunately, goes over my head.
David.
PS I've just joined CM.


Welcome at CM, David.
Many thanks for your appreciation...

Fridger


Top
 Profile  
 
 Post subject:
PostPosted: Tue, 24-01-12, 16:29 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
Here is another pleasing panorama shot, where you can see red glows in various places

Image

In this somewhat magnified portion of the rendering of Orion in Celestia(.Sci), you can see how well the alignment of Nick Risinger's MilkyWay and Celestia's stars actually works:
(although Orion is pretty far away from the center of the MilkyWay (Sgr A*))

Image

Look at the bright star Rigel (β ORI) in the lower right of the image. While the star is from Celestia, the bluish nebulosity around it is from Nick R's MilkyWay photo! Same for the Orion Nebula M42 (blue label!) and the famous Trapezium stars that are ~ centrally embedded:
Image

You can also note well the lower end of Barnard's Loop consisting of red glowing gas (H-alpha line!)

A nice Website about the nebulosities in Orion is here
http://www.allthesky.com/nebulae/orionha.html

From there I remounted two photographs with a H-alpha filter (left) and without (right).
With the Filter you see the red glowing gas (Barnard's loop etc) in great detail

© T. Credner & S. Kohle, AlltheSky.com
Image

Enjoy,
Fridger


Top
 Profile  
 
 Post subject:
PostPosted: Tue, 24-01-12, 20:43 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
Hi all,

today, I received a reply (at last) from Nick Risinger. I had asked him whether he could release a moderately larger size of his MilkyWay image (4096x2048) under a Creative Commons license.

On the one hand, the 4k size would contain considerably more detail than his presently available 3000x1500 map. On the other hand, some older graphics cards might not be able to handle bigger sizes than 4k. So 4096x2048 seems like a good compromise.

Today, he unfortunately wrote to me that he will not make a bigger 4096x2048 sized MilkyWay texture freely available:

Nick Risinger wrote:
...

...I hope you can understand my intentions in keeping it exclusive (at least for the time being). ... Your use of the 3,000x1,500 image is still ok under the creative commons non-commercial attribution license.

Best,

Nick


That's disappointing news, of course.

Fridger


Last edited by t00fri on Tue, 24-01-12, 22:20 GMT, edited 3 times in total.

Top
 Profile  
 
 Post subject:
PostPosted: Tue, 24-01-12, 21:59 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
Just a short while ago, another mail by Nick Risinger arrived. It contained the confirmation of his projection used, as well as a statement about the achieved accuracy in the process of assembling his very many individual frames:

+++++++++++++++++++++++++++
Nick Risinger wrote:
...
you had asked about the projections used in a previous email. The 3,000x1,500 image posted online is a Plate Carrée projection and it should have subpixel accuracy (every plate was assembled via astrometry with distortions eliminated).

+++++++++++++++++++++++++++

This hi-end accuracy matches with my own findings of a practically perfect alignment within Celestia!

Fridger


Top
 Profile  
 
 Post subject:
PostPosted: Wed, 25-01-12, 0:12 GMT 
Offline
User avatar

Joined: Fri, 29-02-08, 17:58 GMT
Posts: 90
Excellent! Thanks for sharing.

Good to see that you got an answer from Monsieur Brunier. And it is unfortunate the decision of Nick, but one needs to understnd the enormous work (and cost) involved in this panorama. Good to know, though, that he was conscious of astrometric precision.

Thanks for the non-flipping also! I tried to do it at first, but I am so bad with models that I couldn't manage, and got stuck with the first version that worked... I didn't have the time, also...

Guillermo


Top
 Profile  
 
 Post subject:
PostPosted: Wed, 25-01-12, 14:40 GMT 
Offline
User avatar

Joined: Fri, 29-02-08, 17:58 GMT
Posts: 90
Fridger,

I see, in the panoramic view with Jupiter, that the photograhic Milky Way is seen, and so is M31, yet not the default Milky Way rendering. Have you implemented what I dreamt, that when you move away from the Sun, the photographic view fades and the blobs rendering appears, until you only see the blobs when you are more than 10 ly away (the radius of the geodesic model? (blobs = your rendering of galaxies, which seems to be much improved over standard Celestia's.)

Guillermo


Top
 Profile  
 
 Post subject:
PostPosted: Wed, 25-01-12, 16:40 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
abramson wrote:
Fridger,

I see, in the panoramic view with Jupiter, that the photograhic Milky Way is seen, and so is M31, yet not the default Milky Way rendering. Have you implemented what I dreamt, that when you move away from the Sun, the photographic view fades and the blobs rendering appears, until you only see the blobs when you are more than 10 ly away (the radius of the geodesic model? (blobs = your rendering of galaxies, which seems to be much improved over standard Celestia's.)

Guillermo


Guillermo,


since I am working on a rather extensive systematic change in terms of "local spheres" (as sketched in http://www.shatters.net/forum/viewtopic ... 7&start=47), no blending in connection with the MW transition is yet implemented in the version I used here.

However, I simply weakened the visibility of the original Milkyway rendering for observers positioned inside our galaxy (i.e. with |x|<R_milky). So in practice, the original, non-photographic rendering is hardly seen / disturbing for |x|< R.

Note that also in my render.cpp code of the standard 1.6.1 version, it is possible to modify the settings accordingly. This just requires rebuilding. The modification does NOT affect the brightness of our galaxy for |x|_observer > R_milky!

The much improved rendering of galaxies and globulars is part of the Celestia.Sci version that I currently use. In full color (SDSS color profile), M 31 looks like so

Image

Another change is that at small Automag zoom (<-> unaided vision) the color saturation may be set to vanish automatically. This simulates the fact that the naked eye only sees galaxies and globulars without colors... There is also a new key that allows to increase color saturation any time according to gusto. To render color of DSOs in a scientifically reproducable manner (color profiles!) and to include multi-wavelength astronomical displays is an important new aspect of Celestia.Sci.

Cheers,
Fridger


Last edited by t00fri on Sat, 28-01-12, 20:27 GMT, edited 2 times in total.

Top
 Profile  
 
 Post subject:
PostPosted: Wed, 25-01-12, 19:21 GMT 
Offline
User avatar

Joined: Fri, 29-02-08, 17:58 GMT
Posts: 90
t00fri wrote:
The much improved rendering of galaxies and globulars is part of the Celestia.Sci version that I currently use.

Yes, these are fantastic improvements.

t00fri wrote:
Note that also in my render.cpp code of the standard 1.6.1 version, it is possible to modify the settings accordingly. This just requires rebuilding. The modification does NOT affect the brightness of our galaxy for |x|_observer > R_milky!

I'll try tweaking it.

Regards,

Guillermo


Top
 Profile  
 
 Post subject:
PostPosted: Wed, 25-01-12, 19:47 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
abramson wrote:
t00fri wrote:
t00fri wrote:
Note that also in my render.cpp code of the standard 1.6.1 version, it is possible to modify the settings accordingly. This just requires rebuilding. The modification does NOT affect the brightness of our galaxy for |x|_observer > R_milky!

I'll try tweaking it.

Regards,

Guillermo


Guillermo,

here are some respective hints:

In src/celengine/render.cpp of Celestia.SVN (or 1.6.1), go to line 10136. It should be the statement

float r = absMag / (float) avgAbsMag;

note my nearby comment:

// obviously, brightness(appMag = absMag) = r and
// brightness(appMag = faintestMag) = 0.2, as desired.

You know that by definition appMag = absMag at a distance of only 10 pc = 32.6 ly from the DSO center. Since galaxies have a much much bigger radius than 32 ly, the quantity r denotes the apparent brightness of the DSO an observer sees from inside.
Aha, that's what you want to decrease...

So all you got to do is to adjust the value of r.
Try e.g.

float r = 0.5f * absMag / (float) avgAbsMag;

The actual value depends on your taste and the settings of your monitor.
People who want to have a brighter Milkyway and don't want to use the photographic MW background should make r bigger e.g. by a factor of 2.

Good luck,
Fridger


Top
 Profile  
 
 Post subject:
PostPosted: Thu, 02-02-12, 18:35 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4514
Location: Hamburg, Germany
Guillermo,

actually, I had forgotten about my own DSO rendering code that I submitted long ago to the Celestia repository ;-)

In render.cpp, it's actually the variable 'enhance' that is best adjusted, if you want a less bright appearance of the default MilkyWay galaxy for an observer located within the galaxy (|x|<R_gal).

Here is the respective comment I wrote in render.cpp:

// The parameter 'enhance' adjusts the DSO brightness as viewed from "inside"
// (e.g. MilkyWay as seen from Earth). It provides an enhanced apparent core
// brightness appMag ~ absMag - enhance. 'enhance' thus serves to uniformly
// enhance the too low sprite luminosity at close distance.

Note that enhance can be negative as well!

You can easily find the place if you search for 'enhance'.

Fridger


Top
 Profile  
 
 Post subject:
PostPosted: Fri, 03-02-12, 1:59 GMT 
Offline
User avatar

Joined: Fri, 29-02-08, 17:58 GMT
Posts: 90
Oh, I see. But the r parameter actually worked well. Thanks.

Render.cpp is well annotated!

Regards,

Guillermo


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 16 posts ]  Go to page 1, 2  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group