Galaxy Rendering

[hank]

Take a look at the "artist's impression" of the Milky Way that accompanies this ESA press release. Apart from the bright core, filaments of dark dust, and scatterred clusters, it seems to me rather similar in general appearance to the rendering seen in Celestia.

- Hank

[Paul]

What, you mean apart from all those very important things you mentioned??? Doesn't seem very similar to me...

It's a bit of a shame that Celestia doesn't have all those things yet... I just hope they're on the list somewhere.

Cheers,
Paul

[Pierebean]

is that normal that we cannot distinguish the different arms(orion one or perseus one) in our galaxy?

[Sky Pilot]

is that normal that we cannot distinguish the different arms(orion one or perseus one) in our galaxy?

Huh? Forgive my ignorance (for it is vast!), but are you suggesting that the Milky Way has two arms? How would we know that?

[selden]

Celestia only uses a generic two-armed "spiral galaxy" blob to represent our galaxy. It does not try to show what is actually known about the Milky Way's overall shape.

My understanding is that the Milky Way is believed to have at least 4 arms, and maybe more.

For more info, see http://www.seds.org/messier/more/mw_arms.html

[maxim]

Shouldn't it have five arms? That's what I remember.

maxim

[selden]

"at least" =or more.

There seem to be indications of 6 and maybe more. The Milky Way's arms seem to be rather fragmentary and not the smooth spirals seen in some galaxies.

The outermost arm seems to contain only gas and no star formation centers yet.

You also have to decide if the Orion Arm (the one containing our Sun) should be counted as a separate arm or not. It seems to be a "spur" -- a short fragmentary arm protrouding either from the next arm out (Perseus Arm) or the one closer to the galactic center (Sagittarius Arm). The evidence seems to be conflicting.

[maxim]

Well, when I came here, I remember that I took a look out of the passenger cabins window, and I think I saw five arms. Our was it M31? Hmm... not shure...

maxim

PS: Selden, I hadn't seen your post when writing, because I didn't refresh my view.

[rthorvald]

Well, when I came here, I remember that I took a look out of the passenger cabins window, and I think I saw five arms

That??s just because you saw double after having one too many Pan-Galactic Gargle Blasters on the flight.

-rthorvald

[Rassilon]

I think it would be quite simple to have celestia render galaxies similar to what I did in my project...Some billboards used as a glow and of course photos of actual galaxies mixed with the current software rendered perlin billboards chris coded...

http://celestialvisions.net/pics/galaxy.jpg

http://celestialvisions.net/pics/galaxy2.jpg

[t00fri]

I think it would be quite simple to have celestia render galaxies similar to what I did in my project...Some billboards used as a glow and of course photos of actual galaxies mixed with the current software rendered perlin billboards chris coded...

http://celestialvisions.net/pics/galaxy.jpg

http://celestialvisions.net/pics/galaxy2.jpg

Ras'

but can you see any method to achieve 'mass-rendering' of galaxies?? Let's say 10000 (NGC-catalog)! That's what we really need to make progress in 'realism'.

Cheers,

Bye Fridger

[maxim]

Well, why not? We have >100000 stars to render. Galaxies would usually fit in a 16x16 pixel raster - exept for the one we're approaching. So it shouldn't be impossible.

maxim

[t00fri]

Well, why not? We have >100000 stars to render. Galaxies would usually fit in a 16x16 pixel raster - exept for the one we're approaching. So it shouldn't be impossible.

maxim

Typically, what catalogs of galaxies contain is some shape outline along with the orientation of that shape as seen from earth...In addition, one usually reads out a classsification symbol corresponding to the galaxy type. The challenge would be to render from that typical base information something that is more realistic than what we have got right now, without 'overdoing' it.

And all this has to be done FAST...

Bye Fridger

[Rassilon]

Fridger,

Yes you simply turn off detailed rendering at great distances...A LOD method which goes from rendering the galaxy model celestia uses, a textured model similar to seldens work. a glow billboard at close range to a single point at great distances....Currently I can render up to several thousand galaxies this way...As you notice in the first picture I posted those smaller dots are a distant galactic cluster comprising of 100+ galaxies...There is currently 10,000 galactic clusters in my project....And this is using an old selection method of rendering....Im still working on a better method of quadtree selection which will increase this number a thousand fold!

Since celestia already uses this method I cannot see it being too difficult adding it in...And the dsc files can be modified to support colorising the billboards etc...that way giving the addon creater the ability to tweak the figures to match current data...

[Evil Dr Ganymede]

I think it would be quite simple to have celestia render galaxies similar to what I did in my project...Some billboards used as a glow and of course photos of actual galaxies mixed with the current software rendered perlin billboards chris coded...

One thing to consider:

I saw the Andromeda galaxy last night through clear skies using a 16" telescope at the observatory down the road... and even through that it looked like a barely visible fuzzy blob with a slightly brighter centre.

It made me wonder - what would a galaxy REALLY look like to human eyes, if you were close enough that it could fill your field of view? We've already had Grant demonstrating that a nebula viewed closeup would be practically invisible to the naked eye. Would the same thing apply to a galaxy too? Would we just be able to see the core clearly and not the arms?

Bear in mind that we're used to thinking that galaxies look like bright cores and arms and dark dust lanes, but that is mostly because we've seen then in long exposure photographs. In most cases, we just can't see the arms and structure with the naked eye, even through telescopes. Would that be the same if we were to see them close up? If so, we won't be seeing views like what Ras linked to here.

(and what's the press release about that hank linked to in the first post on this thread? The link just points to the picture, not to the text that goes with it).

[Toti]

Fridger,

What does this "shape outline" consists of? Could a set of fairly descriptive shape types be derived from it?

[Rassilon]

I think it would be quite simple to have celestia render galaxies similar to what I did in my project...Some billboards used as a glow and of course photos of actual galaxies mixed with the current software rendered perlin billboards chris coded...

One thing to consider:

I saw the Andromeda galaxy last night through clear skies using a 16" telescope at the observatory down the road... and even through that it looked like a barely visible fuzzy blob with a slightly brighter centre.

It made me wonder - what would a galaxy REALLY look like to human eyes, if you were close enough that it could fill your field of view? We've already had Grant demonstrating that a nebula viewed closeup would be practically invisible to the naked eye. Would the same thing apply to a galaxy too? Would we just be able to see the core clearly and not the arms?

Bear in mind that we're used to thinking that galaxies look like bright cores and arms and dark dust lanes, but that is mostly because we've seen then in long exposure photographs. In most cases, we just can't see the arms and structure with the naked eye, even through telescopes. Would that be the same if we were to see them close up? If so, we won't be seeing views like what Ras linked to here.

(and what's the press release about that hank linked to in the first post on this thread? The link just points to the picture, not to the text that goes with it).

I actually think it would be similar to what we see in long exposure photography due to the fact that on the Earth we are able to see galaxies far better without the interference of the full moon or other lights....In space there is hardly any light to interfere, so our eyes would adjust to the absense of lighting similar to long exposure phgtography....Of course the more exposures you take of a galaxy the more pronounced the core becomes...It is true the brightest part of the galaxy is the core and in essence might well be the only thing we see with the naked eye at close range....All of my representations I made in my project is more to see what we normally understand to be familiar with galaxies...and is more pleasing to the eye than an overglorified globular cluster which in reality is what would probably appear....

[t00fri]

Fridger,

What does this "shape outline" consists of? Could a set of fairly descriptive shape types be derived from it?

Toti,

here comes a typical small illustration from the NGC catalog:

Code: Select all

NGC  C D X S P CON    Rect (2000) Decl     Bmag   Vmag   SB     a      b   PA Type        PGC   ID1             ID2             ID3             Remarks                       
--------------------------------------------------------------------------------------------------------------------------------------------------------------
   1   * 1 1 1 PEG  00 07 15.9 +27 42 32   13.6   12.8  13.4   1.70   1.2 120 Sb           564  UGC 57          MCG 4-1-25      ZWG 477.54      ZWG 478.26                     
   1     2                                                                                      IRAS00047+2725  KCPG 2A                                                         
   2   *   1 1 PEG  00 07 17.1 +27 40 43   15.0   14.2  13.5   1.00   0.6 112 Sab          567  UGC 59          MCG 4-1-26      ZWG 477.55      ZWG 478.27, KCPG 2B             
   3   *   1 1 PSC  00 07 16.8 +08 18 05   14.4   13.4  12.8   1.10   0.6 111 S0           565  UGC 58       


where the columns above have the following entries:

Code: Select all

 Field    Explanation
NGC, IC    NGC- or IC-number. Some objects have an extension letter (A,B,C,...).
C    Components. If two or more different objects use the same NGC- or IC-number, these are numbered as components (1,2,3,...). This is also applied for objects associated with the main entry.
D    Dreyer Object. A * marks, that this is the object which is ment in Dreyer's catalogue (many objects with extension letters are not in the original NGC/IC).
X    Second Line. If more then one line is used for an object (if the space for identifications is not sufficient) the lines are counted (1,2) here.
S    Status. Status of the identification (see Tab. 2)
P    Precision. Flag for high precision position
CON    Constellation
alpha, delta     Right ascension, declination. Equinox J2000.0
Bmag    Photographic (blue) magnitude
Vmag    Visual magnitude
SB    Surface brightness (mag/arcmin2)
a, b    Larger/smaller diameter ('). If only one value (a) is given, this refers to the maximum size.
PA    Position angle (?°). The position angle is only relevant for galaxies, and given if possible.
Type    Type of object
PGC    PGC-number. Object listed in the Catalog of Principle Galaxies [22]
ID1, ID2, ID3    Identifications
Remarks    Additional information


So, imagine from entries a,b, the larger/smaller diameters as
given along with their position angle. The type entry provides
information to which of the standardized galaxy classes the
object belongs. Here is a schematic illustration:




I do believe that one might indeed achieve a substantially
improved automatic, generic rendering of galaxies
from such basic data.

It will be far from perfect, but can well contain valuable
quantitative information about size, magnitude, orientation
and galaxy type in a decent graphical display.


Bye Fridger

[Evil Dr Ganymede]

It is true the brightest part of the galaxy is the core and in essence might well be the only thing we see with the naked eye at close range....All of my representations I made in my project is more to see what we normally understand to be familiar with galaxies...and is more pleasing to the eye than an overglorified globular cluster which in reality is what would probably appear....

Therein lies the rub... do we want Celestia to show what galaxies would look like to the naked eye, or do we want it to show what they would look like with a long exposure image?

I think the former would be the ideal, but there's ample justification for the latter - after all, nebulae aren't realistic either, and we don't see the planets as they would to the naked eye because textures have been enhanced in various ways and we don't account for the attenuation of sunlight with distance from the sun.

To be honest, I think the image at the top of the thread is a good comprimise - we can see the structure in the arms, but they're quite dim, and the core is brighter than the arms.

[maxim]

The challenge would be to render from that typical base information something that is more realistic than what we have got right now, without 'overdoing' it.

And all this has to be done FAST...

Well, at first 10000 Galaxies with *individual* 16x16 'far distance' textures would need *less* space than a single 2k texture.
Second, using a culling algorithm that removes invisible instances before rendering would remove most of them (as Evil Dr. pointed out on the andromeda example). Lets say - generously - we'll get 200 remaining galaxies, so the needed pixels are far less than in a 0.5k texture. I'm not sure what size the default galaxy texture is, but I think is about that size, and will have to be processed once for EVERY single instance to be drawn.

So, in general, there is A LOT of potential rendering speed to be gained, while, at the same time, galaxy rendering is considerably improved. In special cases - when we zoom onto a group of galaxies, so that they will show up in 'hires' textures - we'll loose of course our gained benefits, but only to end at the present general situation.

A first step would be to define the format, file storage space, naming and general handling (existence, LOD switch point/distance, ...) of such very lowres galaxy textures.

maxim