New sprites based nebulae

Post requests, images, descriptions and reports about work in progress here.
Avatar
Topic author
Cham M
Posts: 4324
Joined: 14.01.2004
Age: 60
With us: 20 years 10 months
Location: Montreal

Re: New sprites based nebulae

Post #101by Cham » 03.09.2011, 23:19

I've found a nice set of parameters for the lobes :

lobes1.jpg
lobes2.jpg
lobes3.jpg


The view from inside is fantastic ! :D
"Well! I've often seen a cat without a grin", thought Alice; "but a grin without a cat! It's the most curious thing I ever saw in all my life!"

Avatar
Topic author
Cham M
Posts: 4324
Joined: 14.01.2004
Age: 60
With us: 20 years 10 months
Location: Montreal

Re: New sprites based nebulae

Post #102by Cham » 04.09.2011, 20:58

More work on accretion disks.

Here's a plasma torus in the magnetic field of a pulsar (the field lines are desactivated with a script, to better show the plasma). It's the equivalent of Van Allen belts in Earth's magnetic field. It's so pretty, with and without the field lines ! :)

pulsar1.jpg
pulsar2.jpg
pulsar3.jpg


Note : I'll probably publish a huge pack of sprite-based models soon : shells, jets, lobes and disks of various shapes and colors. About 100 models.
"Well! I've often seen a cat without a grin", thought Alice; "but a grin without a cat! It's the most curious thing I ever saw in all my life!"

Avatar
Topic author
Cham M
Posts: 4324
Joined: 14.01.2004
Age: 60
With us: 20 years 10 months
Location: Montreal

Re: New sprites based nebulae

Post #103by Cham » 04.09.2011, 22:23

Dejection models (for pulsars).

dejections2.jpg


I'm not sure I'll go further in that direction (static models ?). Any opinion ?

EDIT : Hmm, pretty artistic pleasure :

dej2.jpg
dej5.jpg
"Well! I've often seen a cat without a grin", thought Alice; "but a grin without a cat! It's the most curious thing I ever saw in all my life!"

Avatar
Fenerit M
Posts: 1880
Joined: 26.03.2007
Age: 17
With us: 17 years 8 months
Location: Thyrrenian sea

Re: New sprites based nebulae

Post #104by Fenerit » 05.09.2011, 01:48

The motion of the electrons along the forcefield lines?
Never at rest.
Massimo

Avatar
Fenerit M
Posts: 1880
Joined: 26.03.2007
Age: 17
With us: 17 years 8 months
Location: Thyrrenian sea

Re: New sprites based nebulae

Post #105by Fenerit » 05.09.2011, 02:16

Goofy wrote:...
Hi Fenerit.
Thank you for your kind words, but if it looks nice is due to your precious "blend add" addition.
...
Goofy :D

I've just suggested the use of, such "addition" is not mine but of developer; now I do not remember where on the forum was discussed as Ideas&News, but it is documented as CMOD propriety within the "changelog.txt" file. The "changelog.txt" file is part of the Celestia's distribution.
Never at rest.
Massimo

chris
Site Admin
Posts: 4211
Joined: 28.01.2002
With us: 22 years 10 months
Location: Seattle, Washington, USA

Re: New sprites based nebulae

Post #106by chris » 09.09.2011, 14:46

Regarding blend modes...

There are three options available in the cmod format: normal, blend and premultiplied, with 'normal' the default. They affect how the geometry--triangles, lines, points or sprites--is blended with the background. For point sprites, each sprite has a color and an alpha value that are multiplied by the sprite texture color and alpha values to produce the fragment color/alpha (Fr Fg Fb) and Fa. The alpha value can be thought of as the opacity, though as you'll see, it's not always treated exactly that way.

Add is the simplest mode. If the background color is (Br Bg Bb), the final color will be:
Fa * (Fr Fg Fb) + (Br Bg Bb)

Normal mode causes the background to be obscured; fragments with an alpha of 1 will completely block the background.
Fa * (Fr Fg Fb) + (1 - Fa) (Br Bg Bb)

Finally, premultiplied is somewhat like normal except it omits the multiplication by the fragment alpha. It assumes that any reduction of the color due to transparency was already done when the geometry was created, hence the name premultiplied:
(Fr Fg Fb) + (1 - Fa) (Br Bg Bb)

The advantage of premultiplied blending is that it can simulate either the add or normal blending modes with appropriately chosen alpha and color values. For example, setting alpha to zero gives the same result as additive blending. Cham: I think this is what you were talking about when you said you wanted to enable add individually for particles.

For nebula, add is good for glowing, diffuse gases where absorption isn't too much of a factor. Premultiplied would in principle let you mix emissive and absorbing particles, but there's a catch: premultiplied and normal blending aren't generally commutative. For correct results, it's necessary to sort the particles from back to front and render them in order, which isn't something that Celestia currently does, as sorting a large number of particles can dramatically reduce performance. In fact, this order dependence of blending has been a big headache in realtime 3D graphics for a long time, and there's still no completely satisfactory solution for the problem.

I'm puzzled by Goofy's result with premultiplied blending. Is it possible to share this add-on so that I can test it?

--Chris

danielj
Posts: 1477
Joined: 15.08.2003
With us: 21 years 3 months

Re: New sprites based nebulae

Post #107by danielj » 10.09.2011, 03:15

Well,it only appears near under Open GL 2.0 render path.
Now I can see.But from near,it disappears...
I download it and didn?t see ANYTHING!Is it possible that I don?t have the specs necessary to run it?

Avatar
Fenerit M
Posts: 1880
Joined: 26.03.2007
Age: 17
With us: 17 years 8 months
Location: Thyrrenian sea

Re: New sprites based nebulae

Post #108by Fenerit » 10.09.2011, 12:45

chris wrote:Regarding blend modes...

There are three options available in the cmod format: normal, blend and premultiplied, with 'normal' the default. They affect how the geometry--triangles, lines, points or sprites--is blended with the background. For point sprites, each sprite has a color and an alpha value that are multiplied by the sprite texture color and alpha values to produce the fragment color/alpha (Fr Fg Fb) and Fa. The alpha value can be thought of as the opacity, though as you'll see, it's not always treated exactly that way.

Add is the simplest mode. If the background color is (Br Bg Bb), the final color will be:
Fa * (Fr Fg Fb) + (Br Bg Bb)

Normal mode causes the background to be obscured; fragments with an alpha of 1 will completely block the background.
Fa * (Fr Fg Fb) + (1 - Fa) (Br Bg Bb)

...

Thanks to add more insights on this. BTW, I think to have seen that in bold when the texture's blob is on a transparent canvas. If the material's opacity is not 1 and the "blend add" attribute is set, supposed (Fr Fg Fb) = 0 and Fa = 1 by default (no color f3/f4 attribute assigned), Add and Normal does inverts their behaviours, because the "disruption" of the (Br Bg Bb) from 1 to 0.
Never at rest.
Massimo

Avatar
Goofy
Posts: 280
Joined: 30.08.2011
With us: 13 years 3 months
Location: Italy

Re: New sprites based nebulae

Post #109by Goofy » 14.09.2011, 23:24

chris wrote:Regarding blend modes...There are three options available in the cmod format: normal, blend and premultiplied
"snip"
I'm puzzled by Goofy's result with premultiplied blending. Is it possible to share this add-on so that I can test it?
--Chris
Hello Chris, sorry for late reply, out of town and without Net access.
Here is a sample of what happens in my system:

NO BLENDING
Image

BLEND
Image

BLEND PREMULTIPLIED 8O
Image

The images are all with Render OpenGL 2.0.
In this small zip file (54 KB) you can find the Nebulae Test addon.

http://www.palatinumclub.it/NebulaeTest.rar

Hope this helps to understand what's gotta happening here.
Bye

Goofy :D
"Something is always better than nothing!"
HP Omen 15-DC1040nl- Intel® Core i7 9750H, 2.6/4.5 GHz- 1TB PCIe NVMe M.2 SSD+ 1TB SATA 6 SSD- 32GB SDRAM DDR4 2666 MHz- Nvidia GeForce GTX 1660 Ti 6 GB-WIN 11 PRO


Return to “Add-on development”