Celestia Forums

It is probably better to continue this discussion in the textures forum:
http://celestiaproject.net/forum/viewtopic.php?t=8540

I am working on a high resolution Saturn map. It can??t BE Saturn as i
haven??t got enough data for it, but my aim is to make a reasonably
realistic representation of Saturn, if not entirely scientifically accurate.

I use bits and pieces from recent Cassini images, and fill in the gaps
with imagination and low resolution where there are no good images
(mostly the nothern hemisphere).

The southern hemisphere isn??t too difficult, but i need help in finding the
right color tones and on how much detail to let through: i am very unsure
how much of these cloudscapes that ought to be visible. Here??s a test
(this is a rough draft, there is much detail to attend to that isn??t addressed
yet here):

Top left: Bj??rn Johnson??s original.
Beside it: the same one, with added noise to get some structure to work from
Next: part of a mosaic of Cassini images, as near the area as i can manage
Last: a blending of the "noise map" and the Cassini mosaic, together with some blurring filters and coloring.

Below: a series of tests to determine "true colors" and visible detail.


... These images are downscaled 2k from 8k. The end result will be a 4k
map. The last test is what i feel will be most realistic, but i don??t know
enough about the Ciclops imaging techniques to decide.

Opinions, please?

- rthorvald

Back in 1998, people at the Space Telescope Science Instute (working for the Hubble Heritage project) made an attempt to create an image of Saturn in "true color" . You might take a look at http://heritage.stsci.edu/1998/29/caption.html
and
http://heritage.stsci.edu/1998/29/big.html
to see if it helps.

p.s. I think the last one is the best, too.

selden wrote:I think the last one is the best, too.

Thank you, Selden. This variant looks like this in Celestia
(left - default Celestia texture on the right):


The colors isn??t very far off the picture you mention either, with no corrections
whatsoever. It is a nice reality-check. But what of the visibility of cloud details...
The question is how much contrast to introduce. Does this seem right?

- rthorvald

Here??s another shot: only the the middle bands in the south
is applied.



-rthorvald

Runar,

I have been thinking a number of times to address a higher
resolution Saturn texture. Great that you are now getting to
it! I suppose you did some research as to the visibility of
these cloud structures in visible light!? I seem to recall that
virtually all Saturn cloud photos are NOT in visible light since
there onde doesn't see much at the distances/resolutions you
are considering.

Bye Fridger

t00fri wrote:I suppose you did some research as to the visibility of
these cloud structures in visible light!? I seem to recall that
virtually all Saturn cloud photos are NOT in visible light since
there onde doesn't see much at the distances/resolutions you
are considering.

You are right: almost all the images shows these features in the infrared.
However, looking at "visible light" images, and comparing to Bj?¶rns map,
it is obvious that Saturn isn??t just a featureless fog in visible light: some
very large features can be made out. The details of these obviously have
texture, but on what level..

That is why i am posting here so early in the process (which i don??t
really like): to get some thoughts and input from others that may know
more than i about how to resolve infrared to visible light.

I have a very hard time determining exactly what ought to be visible.

I am also considering to fix a virtual texture so that one must be *very*
close before the fog resolves to distinct features. But this is far ahead...

As you know, because there arent enough image data, or information
on the positioning, this won??t become more real than Bj?¶rn Jonssons
map: the mosaic i have compiled is touch-and-feel and have huge holes.
But i hope be able to make it more realistic.

- rthorvald

Runar, then, I suppose your third color test was a good compromise...

I don't know about realism either, but it definitely looks a lot more interesting with that kind of detail.
Personally I tried to make "my" Saturn a little more interesting by adding a very faint specular effect. In part to compensate for the lack of support of the haze for ATI cards.
If you want to give it a try...

SpecularColor [ 0.15 0.1 0.06 ]
SpecularPower 1.9

(I'm also trying something similar, yet more apparent with Venus. Its Atmosphere being so flat I was hardly caring about it even though underneath the clouds I placed that excellent VT by John VanVliet.
I'm Thinking about disabling the atmosphere halo completely (since without the haze it would look particularly bad) and use the cloudmap with some specular effect added as the default texture for Venus. Then set a surface map as an alternate map).


Anyway, in my opinion "test 3" image is a good compromise between richer detail from close range, and not too much noise from afar.

Giorgio wrote:I tried to make "my" Saturn a little more interesting by adding a very faint specular effect. In part to compensate for the lack of support of the haze for ATI cards.
If you want to give it a try...

SpecularColor [ 0.15 0.1 0.06 ]
SpecularPower 1.9

I will try that...

Giorgio wrote:in my opinion "test 3" image is a good compromise between richer detail from close range, and not too much noise from afar.

Thank you for your interesting reply. It seems i??m going to do a lot more
testing around #3 and #4. But first i will need to define the nothern
hemsiphere better. Looking at Fridgers "Blue hat" together with these:
http://ciclops.org/media/ir/2005/1474_3662_1.jpg
http://ciclops.org/media/ir/2005/1112_4364_1.jpg
... There are work to be done on Bj?¶rns map before i can proceed...

Re Venus: i have thought about this for some time, too: as with Saturn,
there must be more to see of the clouds than we are able to discern
today. It will be exiting to hear from Venus Express...

- rthorvald

Runar,

since almost precisely one year, my Saturn has a blue head
http://www.celestiaproject.net/forum/viewtopic ... aturn+blue



and a spec-illumination since fare more than a year



Just as a further suggestion for your project...

Bye Fridger

t00fri wrote:since almost precisely one year, my Saturn has a blue head and a spec-illumination since fare more than a year

Yes, i have it: it is quite striking.
I am working on the colors, to get the "blue hat" in my one too.
But before having anything to show, i have some texturing to do.

There??s very little bacground material to use for that part of the map, so
not an easy task - probably must resort to working in some subdued,
inconspicous "noise" from different regions.

Haven??t gotten to working on the ssc definitions yet; there is so much to
do on the graphic for now!

Comparing the existing maps, yours are much more saturated overall, in
addition to the blue tint... Why did you choose this?

- rthorvald

Hi Fridger, I remember the thread about your Saturn texture and specular settings. As a matter of fact, I use your "blue hat" Saturn as the high-res setting in Celestia. It looks very good (definitely better than the standard texture) when the rings cast their shadows on the northern hemisphere. But, I'm afraid it may not be very realistic in any other condition, when, from the pictures I've seen, it appears like Saturn has more or less symmetrical colors for the two halves.

Also I tried your specular settings last year, but it made the planet look like plastic, in my opinion.

I tried several settings, untill I was quite satisfied with the ones I posted before. They look like this: (image is pretty wide but just 39Kb heavy)



Poor ATI 9550, can't handle haze, but still, since I'm short sighted, I noticed watching this picture on the screen with my old heavy glasses at about 30?° to the right of the screen ...thanks to optics a nice blue haze appears on the right side of Saturn
Pathetic, I guess

Status: some screenshots from a 4k PNG test...




These are a bit more saturated than the previous ones, with a "hint" of the Blue Hat. I have also added Fridgers specular values, and enhanced the clouds with a very discreet 2k bumpmap.
What do you think?

-rthorvald

I think they look great, but...

Could you be persuaded to use thumbnails in your post that link to the large images? Large pictures cause very long load times, especially for people with slow or even dialup network connections. Thumbnails will let the topic load quickly for everyone but still provide access to the high resolution pictures.

The code to do this is

selden wrote:I think they look great, but...

Could you be persuaded to use thumbnails in your post that link to the large images? Large pictures cause very long load times, especially for people with slow or even dialup network connections. Thumbnails will let the topic load quickly for everyone but still provide access to the high resolution pictures.

The code to do this is

Code: Select all

[url=big-picture.jpg][img]small-picture.jpg[/img][/url]


I am on Runar's side. I will NOT do this with my images.
Otherwise I'll stop publishing ANY on this site! There might
be alternative options for my (and Runar's) threads, in
fact ...


I don't think the downloading time for such jpegs is THAT
long...


Bye Fridger

t00fri wrote:I am on Runar's side. I will NOT do this with my images.
Otherwise I'll stop publishing ANY on this site! There might
be alternative options for my (and Runar's) threads, in
fact ...


I don't think the downloading time for such jpegs is THAT
long...

Fridger, definitively, the loading time is not so long, but what Selden mean (I suppose) is that having to reload the same images (already seen) each time you visit a thread can become a pain for dialup users...

Runar, the work seems very subtle... personally I would like to see a small band (full res) of the 4k map to judge better...

selden wrote:Could you be persuaded to use thumbnails in your post that link to the large images? Large pictures cause very long load times, especially for people with slow or even dialup network connections. Thumbnails will let the topic load quickly for everyone but still provide access to the high resolution pictures

I am nut sure how to respond to this, Selden... I certainly don??t want to make things difficult for anyone, and if this had been a regular discussion on any topic BUT graphics work, there??d be no question!

But... This is ALL about graphics. That is the entire purpose. I knew it would become huge, so i labelled the thread with a warning. And i use my own bandwidth, so it does not slow down the forum... Isn??t that enough?

As i see it, the only reason for reading this thread is to look at the graphics and comment. If you have seen it once, your browser caches it for next time, so the regular user loads the picture only one time...

Please consider. I??ll wait for your reply before posting more images.

- rthorvald

ElChristou wrote:the work seems very subtle... personally I would like to see a small band (full res) of the 4k map to judge better...

Here:
http://runar.thorvaldsen.net/celestia/temp/satest3.jpg

- rthorvald

rthorvald wrote:Here:
http://runar.thorvaldsen.net/celestia/temp/satest3.jpg

Do I see some noise here and there? seems ok to me because OpenGL smooth a bit this kind of features and make a texture more "alive"...

Here is my real point I want to make:

Again, I tend to observe an increasing "flatness" in the posts that appeared in the forum, recently. Very little substantial material anymore.

1) Hey my display is full of "red sprinkles", can you help?
2) Klapperschlangen
3) ...

So I think we should be as receptive and grateful of any more "in depth" contributions with REAL SUBSTANCE"!

That's why I am on Runar's side .

You may have noticed that "t00fri" logins are getting considerably more rare these days:

1) I am bored to death about what I can read here (so seem to be quite a few others whose presence I value!)

2) I am working my head off for Celestia, instead

Bye Fridger