Galaxy Templates via STANDARD .BMP Format, DOWNLOAD!

[t00fri]

Since Toti and I have discussed our different points of view among us to a point where further progress seemed improbable, I suggest we do not start this over again here.

There is lots of interesting work to be done, instead...

Toti, if you indeed think that you can realize a galaxy (viz DSO) display

-- that is ~ as fast and storage economic as the present one,
-- allows to extract great galaxy displays from standard scientific catalog values only
-- incorporates great resolution + depth sorting +...at no prize in performance
-- all your advocated customization and editing ability

I'd say, realize this in your Sandbox at home and let us know the result by all means! Everyone will surely be happy to discuss and test them on such a more concrete basis! That's the way Celestia development has always been operating.

That's also the answer to your "question" above:

And how could I show a single benchmark of a new massive reimplementation without big modifications to the code remains an ontologic mistery to me... (that I DO NOT want to discuss again)

You can always do as many modifications of the CVS code in your own Sandbox as you want....It's the results that count.

I am just not ready to have our quite successful galaxy code reshuffled entirely before it is really obvious what we get in exchange! Notably, such an effort does NOT seem worthwhile for me, if at the end of the day there would at best be 10-15 galaxies, where we have enough scientific information to incoporate all this new fancyness!

I think sometimes, I also know what I am talking about

Bye Fridger

[Toti]

That's also the answer to your "question" above:

And how could I show a single benchmark of a new massive reimplementation without big modifications to the code remains an ontologic mistery to me... (that I DO NOT want to discuss again)

You can always do as many modifications of the CVS code in your own Sandbox as you want....It's the results that count.

I am just not ready to have our quite successful galaxy code reshuffled entirely before it is really obvious what we get in exchange! Notably, such an effort does NOT seem worthwhile for me, if at the end of the day there would at best be 10-15 galaxies, where we have enough scientific information to incoporate all this new fancyness!

That's obvious now that I have full CVS access and Celestia is progressing again, but at the time I am referring there the workflow was much more limited, remember? Also there was no irreversible modifications involved from my side nor any deletion of your own sandbox.

Toti, if you indeed think that you can realize a galaxy (viz DSO) display

-- that is ~ as fast and storage economic as the present one,
-- allows to extract great galaxy displays from standard scientific catalog values only
-- incorporates great resolution + depth sorting +...at no prize in performance
-- all your advocated customization and editing ability

I'll probably take a second round on all this when I get new video hardware, with higher fill rates and a load of more advanced features. A TNT2 with AGP 1x is simply too limited for this stuff.

I'll stop here for now.

[ElChristou]

...please google a little bit on the subject: there are real time animations out there made with millions of blobs...

Toti, I want to read more on the topic, would you be kind enought to point me on some specific terms (words, expressions) for googling? (I don't know all the specific vocabulary used for this kind of work).

[fsgregs]

Fridger:

I was looking through your creator's files on Celestia and encountered this image of M74. Is this based on your new code? It is REALLY GOOD!





Frank

[t00fri]

Fridger:

I was looking through your creator's files on Celestia and encountered this image of M74. Is this based on your new code? It is REALLY GOOD!

Frank

Frank,

the new PNG template format gives in principle the same quality if the old .pts templates are /converted/ to PNG. However instead of ~300KB storage request for a .pts template, the new /equivalent/ PNG template
only needs ~ 5KB!

++++++++++++++
The greatest new aspect is, however, the fact that PNG is a standard lossless format that may be edited directly with EVERY image manipulation software, like GIMP, PS, ...
++++++++++++++

So since > 1 week I am working intensively with ElChristou on an entirely new PNG template set! For sure, there are meanwhile in excess of one hundred emails exchanged between Paraquay and Hamburg/Germany

ElChristou has invented a really neat and systematic multi-layer design strategy for general templates, that we are currently realizing. Clearly the next step will be the extended Hubble classification with many more templates that we can easily afford now, from the point of view of storage consumption.

Since all this is nevertheless quite tedious, it will still take a little while before we are both content with our results and I can commit the whole stuff to CVS.

The image (m74) you were referring to above looks considerably more fancy meanwhile. Like so:

Bye Fridger

[Cham]

This looks VERY impressive ! I can't wait to see this live in Celestia !

[snabald]

That is AMAZING!!!!

Will this be in the next version?

[bdm]

I was searching the net for something else and I found an interesting link that's relevant to this discussion. I thought I would post it here in case it was useful.

http://www.uni-weimar.de/~hildebr2/download/thesis.pdf

[buggs_moran]

I was searching the net for something else and I found an interesting link that's relevant to this discussion. I thought I would post it here in case it was useful.

http://www.uni-weimar.de/~hildebr2/download/thesis.pdf

Very interesting thesis. Are we seeing the future possibilites of Celestia? He lost me on a couple of points (mostly because my college level thinking is diminishing with each passing decade.) However, over all there are some really amazing possibilities here. The planetary nebulae were wonderful http://www.mpi-sb.mpg.de/~magnor/civr.html. I like Fridger and Toti's galaxies better though.

[t00fri]

I was searching the net for something else and I found an interesting link that's relevant to this discussion. I thought I would post it here in case it was useful.

http://www.uni-weimar.de/~hildebr2/download/thesis.pdf

Very interesting thesis. Are we seeing the future possibilites of Celestia? He lost me on a couple of points (mostly because my college level thinking is diminishing with each passing decade.) However, over all there are some really amazing possibilities here. The planetary nebulae were wonderful http://www.mpi-sb.mpg.de/~magnor/civr.html. I like Fridger and Toti's galaxies better though.

As to galaxies, I used the same formulae to display them in the correct orientation starting from face-on shapes. Yet these formulae are even found in the professional catalog explanations...

What is not clear to me yet is whether there is something in the nebulae rendering stuff that we might adapt for high-speed rendering in Celestia. I also prefer our galaxy rendering based on sprites by far...

Buggs: this URL that you found seems definitely usable
http://www.mpi-sb.mpg.de/~magnor/civr.html
I'll definitely get back to this, because rendering planetary nebulae in Celestia gave me quite a bit of a headache up to now...

As a Diplome thesis its certainly not a boring subject.

Bye Fridger

[t00fri]

Hi all,

reading these papers on "constrained inverse volume rendering" of (bipolar) planetaries clearly indicates the path to take for planetary nebula rendering in Celestia:

What the authors realized is the /axial/ symmetry of planetary nebulae. This implies that we can render the /full/ 3d structure in terms of a (rotated) 2d template density function that one takes from suitable photographs like in case of galaxies! The extremely varied appearances of planetaries if viewed from earth then just arise due to different orientations these objects subtend! You can see these effects in their fascinating movies.

+++++++++++++++++
It seems that my headache is definitely over in this respect. Thanks again to bdm and buggs for locating these important sources of information!!
+++++++++++++++++

Bye Fridger

[jll]

I do not know if that can be useful, but in the various documents which I could read on the 3D modeling of nebulas, except the excellent documents of Dr. Magnor you mention, I also found these sites (that the state of my scientific knowledge did not allow me to exploit ).

A Planetary Nebula Classification
http://www.lam.oamp.fr/trung/page1.html

A lot of documents from Max Planck Institut (including civr)
http://www.mpi-inf.mpg.de/~lintu/projects/aat.html (see the references at the end of the page)
http://www.mpi-inf.mpg.de/departments/irg3/data/rnv.html (with some short C code)
http://www.mpi-sb.mpg.de/~magnor/rnv.html (reflection nebula)

An older one
http://www.huginn.com/knuth/pne.html

Bye
JLL

[buggs_moran]

Fridger,

Could you use something similar to your sprite based galactic objects for emission/reflection nebulae? Individual nebulae obviously cannot be classed by shape like galaxies and therefore this approach would not work on a one by one basis. Instead, what if we did an overlay sprite based on a local intersteller density map for different elements and dust. This way there would be no individual nebulae, just whole regions of gas and dust based on a local map.

Can sprites interact with the luminosity of stars in Celestia like planets do? If so, couldn't a region of dust be "lit" up by a nearby star? If the star is inside, the "gas/dust" would redden (emission), if outside, the "gas/dust" could turn slightly blue (reflection). By relying on stellar temperature, distance and luminosity we could eliminate the need for colorizing the molecular clouds individually. Secondly, would it be feasible for the regions in Celestia to extinguish (redden) starlight as we looked through them?

Finally, could individual nebulae be sprite mapped based on radio maps? Possibly just for diffuse nebulae, the Messiers for instance (to keep the numbers low). The density from the map could correlate to the thickness of the cloud in Celestia. Perfect? No, but closer than a png pasted to a billboard cmod. Not that those aren't pretty. I really don't know what kind of undertaking all of this would be. I would be happy to do my part by collecting data and getting permission to use research materials if needed.

I have found some sites on the subject for further reading if anyone is interested in this.

For the galaxy as a whole
http://www.ldps.ws/Mirror/Universe/milkyway.html

Local Bubble info
http://www.berkeley.edu/news/media/releases/2003/05/29_space.shtml
http://geophysics.tau.ac.il/personal/oferya/documents/The%20Galactic%20Environment%20final%20appendix.pdf
http://www.solstation.com/x-objects/chimney.htm

For individual diffuse nebula the maps on this page may be of some use.
http://www.ldps.ws/Mirror/Universe/m42.html

[t00fri]

Excellent, guys!

many thanks for digging out all these interesting references! At this point it's simply too early for me to make statements about the possibly best options for rendering such nebulae within the framework of the MPI papers. Certainly sprites are the first thing to explore.
Certainly one can hope to make a lot of progress.

We should avoid mixing in now different sorts of nebulae! Diffuse nebulae are still problematic, since there is no simple classification scheme. But the great new aspects refer to (bipolar) /planetary/ nebulae.

Bye Fridger

[t00fri]

Buggs,

I am not sure whether you got the main message behind those MPI papers: In the past the observation of (bipolar) planetary nebulae from earth has confronted us with a GREAT variety of shapes and colors etc. The message is that there is an axial symmetry in (bipolar) planetaries that allows us to reduce the full 3d description to a (rotated) 2d profile that essentially has a lot of UNIVERSAL features. The great variety of views arises mainly from the different prespectives/ orientations we observe from earth!!

That fits perfectly into my mass rendering approach for deepsky objects that I have been advocating from the beginning.

Bye Fridger

[tech2000]

Hi Fridger. I hope this Celestial Matters is what I think it is... Nice done.

[buggs_moran]

Buggs,

I am not sure whether you got the main message behind those MPI papers: In the past the observation of (bipolar) planetary nebulae from earth has confronted us with a GREAT variety of shapes and colors etc. The message is that there is an axial symmetry in (bipolar) planetaries that allows us to reduce the full 3d description to a (rotated) 2d profile that essentially has a lot of UNIVERSAL features. The great variety of views arises mainly from the different prespectives/ orientations we observe from earth!!

That fits perfectly into my mass rendering approach for deepsky objects that I have been advocating from the beginning.

Bye Fridger

I understand the basic premises, that's why I included the link a few posts back. After all, further possibilities of a high quantity of objects being rendered at a low computer cost (memory and processor) has been your forte in Celestia. While I was looking at the videos I had a whole Gabriel's horn flashback from college Calculus. For those who don't remember, take y= 1\x, constrain it to values greater than or equal to one, and spin it around the x axis, et voila, a horn (with finite volume and infinite surface area to boot).

My last post above, dealing with the ISM, was on a totally different topic and I shouldn't have necessarily included it here. I just thought I would bring it up while we were on the subject of nebulae.

[buggs_moran]

Decided to separate this from my other post.

More links on simulated planetary nebulae
http://www.strw.leidenuniv.nl/~icke/html/VincentScience.html#anchor20553087
http://www.huginn.com/knuth/pne.html

Classification list for PN
http://www.blackskies.org/class1.htm
and an Excel file listing 1143 Planetary Nebulae with type, RA, Dec, Distance, Size, Mag and Surface Brightness. It would have to be pared down to wipe out some PN's without distance data.
http://www.blackskies.org/PN_Files/SECGPN_V6DS.xls

[buggs_moran]

Really nice movie of movement in a bipolar nebula at APOD today:
http://antwrp.gsfc.nasa.gov/apod/ap070618.html

Thought I'd put that here to hopefully reignite some thought on these nebulae in Celestia...

[makc]

Check out this flash signature I did back in 2005, maybe it will give you some ideas The "galaxy" is generated, but you can do the same with bitmap texture very easy.

Edit: in case it is not obvious how that works, I thought I'd add few words about it. The galaxy "image" is sliced into ~100 images of the same size stacked over e.o., so that "pixels" near the top go to deepest image in the stack, next scanline goes into image #1, and so on, and then images in stack are simply shifted vertically to simulate rotation. To simulate volume near the bulge, I just added random noise to the number of image where pixels go.