New accretion disk method (WORK IN PROGRESS)

[Cham]

Another one:

[Cham]

A new black hole model, all sprites based (fully volumetric effects) :

[Hungry4info]

That is amazing

[daerokos]

Ok, here's the ring used around our Sun (just as a test) :

EDIT : And the sprites may even be used as special markers, or even billboards floating always face-on in space. Sadly, the sprite size appears to be limited (can't be bigger than on this picture).

billboard.jpg

[/quote]


so i downloaded your sprites test.zip and i changed the star name in the ssc to the name of my star. i was wondering how you would make it so that the ring would orbit further away from the star. i tried changing the semi-major axis of the #2 thing, but it just extended an orbit with nothing around it. the ring sprites stayed the same distance from Odin. is it impossible?

[Celestial234]

cham how to you make it?

[Cham]

Spectacular view from Celestia ! (really need some thick sunglasses here)

redhot5.jpg

redhot4.jpg

[Cham]

I'm using plain opaque non-blurry disks to simulate big rocks. The dust atmosphere is so strong in Celestia (especially while the ring is rotating), I'm tempted to cough...

dust4.jpg

dust5.jpg

dust6.jpg

[fsgregs]

Cham:

NOW THAT is what we're talking about!!!!!!

beautiful stuff. Very realistic, particularly for simulating a new solar system in the making. Will you be releasing any formal addons using it?

Frank

[Cham]

Frank,

yes, I intend to release that dust ring with its pebbles as a small addon for Fomalhaut. The addon would be very simple : just a single planet, using Fenerit's texture, with a dust/dirt ring. The user could modify the addon to adapt it to his particular needs.

Probably in the next few days, so stay tuned...

[Cham]

I must admit that this thing is getting pretty impressive, live in Celestia. I can hear the stormy noise, with all these rocks flying around (of course, there's no sound in space...).

I added a darker component, for more dramatic effect :

storm1.jpg

storm2.jpg

storm3.jpg

[Fenerit]

Martin, take a look whether png like below can help in "coughin"... These are easy to do through 3D particles system.

accr.png

[Cham]

Martin, take a look whether png like below can help in "coughin"... These are easy to do through 3D particles system.

accr.png

No, this can't be of any help, since sprites have a constant orientation relative to the screen.
Non-spherical "blobs" or non-circular shapes are terrible with the sprites technique.

In other words, ALL individual sprite MUST be a circular shape. This is a limitation of the sprites rendering.

If you use any non-circular shape, the global rendering would be horrible, especially considering the orientation of the light source and the motion of the observer.

EDIT : Massimo, can you do the same kind of lunar texture you have done before (in 4k please), but adding volcanic effects ? Lava cracks and rivers ?
Of course, this implies a night texture, and large height normalmap.

[Fenerit]

No, this can't be of any help, since sprites have a constant orientation relative to the screen.
Non-spherical "blobs" or non-circular shapes are terrible with the sprites technique.

Ops! Right, I forgot that.

EDIT : Massimo, can you do the same kind of lunar texture you have done before (in 4k please), but adding volcanic effects ? Lava cracks and rivers ?
Of course, this implies a night texture, and large height normalmap.

I think yes but I do not remember which lunar: Callisto/Ganymede alike or the Enceladus alike? The Enceladus alike destroy the sight in do it. It impress the retina, an ugly experience. Do you mean lava river?

[Hungry4info]

Wow that looks amazing!

[Cham]

I placed the dusty planet around Beta Pictoris, a star well known for having an accretion disk of dust and debris.
The addon is very simple :

2 planets. One fictious dusty mars-like planet (texture made by Fenerit) with the accretion ring and debris, and some exoplanet (not confirmed yet).
There's a simple "generic" dust ring around the star. The martian-like planet texture is missing some lava fields, craters and eruptions, though.
I'm just tired to use the same venus texture as what we see everywhere !

Here's a preview :

beta_pictoris.jpg

The view is beautifull only when the observer moves very close or inside the dust system. Accelerate time by a factor 100 or 1000, for the cinematics !

It is sad that Celestia currently can't drop a shadow on its dust ring system, like what it can do for a standard ring.

Here's a download link to the addon :

http://fsgregs.startlogic.com/Public_Fi ... ctoris.zip (7.8MB zip file)

Do whatever you want with this, but only for non-commercial activities. If you use it for something else in Celestia and publish it, remember to give credits !

Two close-ups of the martian-like dusty world (cough-cough !) :
The darker belt in the background is the generic dust field around beta Pictoris.

bet1.jpg

bet2.jpg

[Goofy]

Hi Cham, thsnks a lot for uploading it.
As I told you in the "Pictures from Celestia" thread I tried it on Saturn, and here the results:
If it will be possible to reduce the particles ring thichness, reduce the dimensions of sprites and making them white, it will give a good idea of Saturn rings.
Cham, can you help me?
Thanks a lot anyway, your wiork is outstanding, as always.
Bye

Goofy

[Cham]

Goofy,

editing a CMOD sprites file is really easy, if you have a spreadsheet app like Excel (I prefer to use NeoOffice).
Here's a small tutorial about sprites data :

Open the file with your favorite text editor. The header is a declaration of several material definitions :

Code: Select all

#celmodel__ascii

material  # invisible material for basic mesh below
  emissive 0 0 0
  diffuse 0 0 0
  opacity 0
end_material

material  # 1 Blobs
  #emissive 0.8 0.2 0.3
  diffuse 0.2 0.15 0.12
  opacity 0.99
  texture0 "disk.png"
end_material


The first material (number 0) is just for the small mesh that is used to center the model correctly in Celestia (see below). It must be invisible, so don't touch this one.

The second material (number 1) is for the mesh number 1 (see below). For full white, change it to this :

Code: Select all

material  # 1 Blobs
  #emissive 0.8 0.2 0.3
  diffuse 1 1 1
  opacity 0.99
  texture0 "disk.png"
end_material


or this, if you what the sprite to glow white :

Code: Select all

material  # 1 Blobs
  emissive 1 1 1
  diffuse 1 1 1
  opacity 0.99
  texture0 "disk.png"
end_material


The sprite opacity should stay at 0.99, since the value 1 (full opacity) gives depth sorting troubles in Celestia. if you want the sprite to be invisible, set this to 0. Usually, transparency is achieved with numbers like 0.25, or 0.5, for example.

Then comes the meshes. Mesh number 0 is used as a reference object to center the model correctly in Celestia :

Code: Select all

mesh  # basic mesh to center the model in Celestia
  vertexdesc position f3 end_vertexdesc

vertices  6
20 0 0
-20 0 0
0 20 0
0 -20 0
0 0 20
0 0 -20

linestrip 0 2
0
1

end_mesh


Don't touch this one. Next come the mesh number 1, which looks like this :

Code: Select all

mesh
  vertexdesc position f3 pointsize f1 end_vertexdesc

vertices  10000
5.63945639   5.62153656   0.23402109   0.01
-8.59911278   -0.28225359   -0.20370865   0.01
2.55810706   -7.09169358   0.34004022   0.01
0.70764691   -8.64986119   0.03110831   0.01
-3.30325399   -8.25897190   0.18711892   0.01
....


The three first columns are the X, Y, Z cartesian coordinates. The fourth column defines the sprite size S.
If you want smaller sprites for Saturn's rings, then change S = 0.01 to a much smaller number, something like 0.0001. But then, the sprites wont be visible at a large distance. You will have to enter the rings to see some of them.

If you want to define a flat distribution of sprites, then change all the Z coordinates to 0, like this :

Code: Select all

mesh
  vertexdesc position f3 pointsize f1 end_vertexdesc

vertices  10000
5.63945639   5.62153656   0.0   0.0001
-8.59911278   -0.28225359   0.0   0.0001
2.55810706   -7.09169358   0.0   0.0001
0.70764691   -8.64986119   0.0   0.0001
-3.30325399   -8.25897190   0.0   0.0001
....


If you want to distribute the sprites on some small thickness, just add some small random numbers, between -0.01 and 0.01, say (you can use the RANDOM function of the spreadsheet app for this) :

Code: Select all

mesh
  vertexdesc position f3 pointsize f1 end_vertexdesc

vertices  10000
5.63945639   5.62153656   0.0000543   0.0001
-8.59911278   -0.28225359   -0.0054545   0.0001
2.55810706   -7.09169358   0.0075576   0.0001
0.70764691   -8.64986119   0.0095643   0.0001
-3.30325399   -8.25897190   -0.006582   0.0001
....


That's it !

However, take note that you'll need MUCH MORE sprites, for planar rings made of very small sprites ! Or else, you wont see anything in the rings, since the sprites are small and very far away from each other. For satisfying planar rings, you need something like 100 000 or even several millions of sprites ! This may have a severe impact on the frame rate ! I'm not sure this is worth it.

[Cham]

Hmmm ! This is more interesting than what I though :

I just made a rough ring distribution on Saturn, with 100 000 small sprites. The frame rate is VERY good, and the ascii CMOD file is "just" 5.7MB. Here's a preview :

saturn.jpg

I'll do more experiments with this...

[Goofy]

Hi Cham, thanks a lot for the detailed cmod files editing help, very appreciated.
I see that my Saturn test interested you to try further experiments, so that you'll show us more results on this matter.
Waiting for them.
Just two suggetsions: the sprite ring size/distance from Saturn should be fine-tuned, actually there is a lot of particles in the D ring, where they shoul be just a little number.
The second is: could it be possible to divide the sprites ring in 3 different rings, at the distance and large as rings A-B-C, and moreover with the corresponding sprite densities?
I mean e.g.: 12000 particles in C ring, 63000 in B ring, and 25000 in A ring (numbers are just an example, I've not searched for albedos and width of the 3 rings)
Bye soon.

Goofy

[Goofy]

More about Saturn rings.
We know that 2- 3 and even more sprites cmods can be in the same position, as you showed us with your nebulae.
My suggestion: why not use 2, 3 or even more cmods in the same position but with different particles sizes and luminosity?
Moreover, their number (=density) should be tuned respect to their size, e.g. 1% size=1, 20% size=0.7 and the remaining 79% size= 0.4 or so.
I think all this could give a lot more realism.
Just my little cent.
Bye

Goofy