Revamping Steve Albers' 4k Mercury texture...

[t00fri]

---------------------------------------------------
EDIT: My Mercury images in this and the following post have been edited.
For explanations see further down:
viewtopic.php?f=5&t=13671&start=7
----------------------------------------------------

Well.

first of all, it is not my intention to criticize Steve Albers' 4k Mercury texture...
http://laps.noaa.gov/albers/sos/sos.html

While it may be OK for some purposes, it looks really ugly under Celestia rendering, as various users have already noted. What makes the issue particularly touchy is that Chris Laurel has committed this texture to SVN for the forthcoming Celestia 1.6.0 release.

Independently of all this, I am underway to do a Mercury texture from scratch, using all existing Messenger imaging along with lots of reprojections via ISIS3 etc. The result looks great so far, but it will be for Celestia.Sci rather than for Celestia proper.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Anyhow, as a large improvement of the visual experience in Celestia proper, I have offered to revamp the recent Mercury texture by Steve Albers considerably, without sacrificing any imaging accuracy ...

Let me give you below a flavor of my results from a single afternoon's work...
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Here are some comparison images. So you can judge for yourself...

• simple cylindrical projection, reduced by a factor ~ 6 :
  
  Steve's Original (click for BIG):
  --------------------------------------
  
  
  My improved texture(click for BIG)
  -------------------------------------------
  
  (download the 4k texture directly from here if you like. But watch out not to take the reduced version, accidentally!)
  
  
• Now some representative Celestia renderings:
  
  Steve Albers' original
  ---------------------------
  NO this is NOT Iapetus
  
  
  
  My revamped texture:
  ---------------------------
  
  
  Steve Albers' original
  ----------------------------
  
  
  My revamped texture:
  ---------------------------
  
  
  Steve Albers' original
  ---------------------------
  
  
  My revamped texture:
  ---------------------------
  
  
  
  
• As usual, my texture is colored by computer with a Mercury color photo from Messenger as a coloration template.
  
  
  
• +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
  I am ready to submit this improved Mercury texture to SVN for Celestia 1.6.0,
  but I am not ready to argue with Steve...
  
  Let me know.
  +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Fridger

[t00fri]

I forgot to illustrate, how much less offending to the eye my limit_of_knowledge region is (where imaging is lacking), compared to Steve's bright white color. As usual, I determine the average color (GIMP) of a texture block in the neighborhood. This color is then used as LOK color. The lack of data is still visible, but the view is nevertheless quite tolerable.



Fridger

[BobHegwood]

Very, VERY nice there Good Doctor...
Many thanks for your efforts on this front.

As always, your textures are nothing short of superb.

[t00fri]

Good you like it, Bob,

happy Easter,
Fridger

[ElChristou]

Nice job! Now ok for a new mercury!

[Reiko]

Looks fantastic! Any idea when the one you are making for celestia sci will be available? I like textures better with the unknown regions filled in instead of having blank areas.

[t00fri]

Looks fantastic! Any idea when the one you are making for celestia sci will be available? I like textures better with the unknown regions filled in instead of having blank areas.

Happy Easter, Reiko,

since I am working all by myself on Celestia.Sci for now, with lots of new features to be incorporated, it will take quite a bit of time before I can go public. Also, I don't want to spend all my spare time in front of a computer

As the acronym is to indicate, Celestia.Sci will focus even more on scientific aspects of Space visualization than does Celestia proper. Hence, my textures and database will certainly not contain any "artistic additions".

Fridger

[t00fri]

Hi all,

while yesterday, I had taken Steve Albers' usual central Meridian convention as well as his Mercury color as a basis for my above display, this required some corrections today after further checkings:

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Note: In my main post at the beginning of this thread, I have edited the required changes! .
Have a look once more to my new/revised images above!
viewtopic.php?f=5&t=13671&start=0
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

• Exceptionally, Steve used Celestia's convention for Mercury, hence the Meridian of my display above was incorrectly rotated by 180 degrees . Here is an official display from NASA/Messenger, with the 0 degree meridian in the texture center as required by Celestia:
  
  
  
  
  
• There have been plenty of studies about the true color of Mercury by the Messenger team. It seems that Mercury is less colored to the naked eye then the enhanced color patches on Steve Albers' texture. Here is Mercury in "best" true color from Messenger, which I have now used throughout as a coloration template:
  
  

Enjoy,
Fridger

[ElChristou]

Is there a normal or bump map in your last shot?

[t00fri]

Is there a normal or bump map in your last shot?

Absolutely not.

You know you could just download the 4k texture above and try it yourself...

Fridger

[chris]

Looks very nice, Fridger. I like your choice of more subtle coloring--is it based on the lower right image in this quartet:

http://photojournal.jpl.nasa.gov/catalog/PIA11364

?

(For anyone interested why its so hard to find 'true color' images of planets, the caption for the images above is worth reading.)

--Chris

[chris]

Looks very nice, Fridger. I like your choice of more subtle coloring--is it based on the lower right image in this quartet:

http://photojournal.jpl.nasa.gov/catalog/PIA11364

?

(For anyone interested why its so hard to find 'true color' images of planets, the caption for the images above is worth reading.)

I will ask Steve whether he is ok with us using a modified version of his texture for Celestia 1.6.0.

--Chris

[t00fri]

Looks very nice, Fridger. I like your choice of more subtle coloring--is it based on the lower right image in this quartet:

http://photojournal.jpl.nasa.gov/catalog/PIA11364

?

Yes, of course, the one on the lower right

I will ask Steve whether he is ok with us using a modified version of his texture for Celestia 1.6.0.

Yes, please... We'll also have to find an adequate reference to both of us for the README...

Fridger

[Reiko]

Looks fantastic! Any idea when the one you are making for celestia sci will be available? I like textures better with the unknown regions filled in instead of having blank areas.

Happy Easter, Reiko,

since I am working all by myself on Celestia.Sci for now, with lots of new features to be incorporated, it will take quite a bit of time before I can go public. Also, I don't want to spend all my spare time in front of a computer

As the acronym is to indicate, Celestia.Sci will focus even more on scientific aspects of Space visualization than does Celestia.proper. Hence, my textures and database will certainly not contain any "artistic additions".

Fridger

Oh I see. I thought you were making one with the gaps filled in. My reading confusion gets the better of me again.

Happy Easter!

[BobHegwood]

As the acronym is to indicate, Celestia.Sci will focus even more on scientific aspects of Space visualization than does Celestia.proper. Hence, my textures and database will certainly not contain any "artistic additions".

Fridger

To this, I can only say "Hear, hear." Great news for the realists of us in the world
Good Doctor. If I can ever help in any way, please don't hesitate to ask.

You should know how very much I prefer the realistic approach by now.

Many thanks, Brain-Dead

[ElChristou]

Yep, this 4k is quite an improvement over the old map. Hope Steve A. will be ok to let us implementing this map in the default package...

[t00fri]

As the acronym is to indicate, Celestia.Sci will focus even more on scientific aspects of Space visualization than does Celestia.proper. Hence, my textures and database will certainly not contain any "artistic additions".

Fridger

To this, I can only say "Hear, hear." Great news for the realists of us in the world
Good Doctor. If I can ever help in any way, please don't hesitate to ask.

You should know how very much I prefer the realistic approach by now.

Many thanks, Brain-Dead

Bob,

many thanks for your interest and support! Help is always welcome (once the "baby" has learned to stand on its own feet ).

Cheers,
Fridger

[RVS]

Hi!
Some doubts about Mercury’s color.
1. Honestly, the caption for picture PIA11364 is incomprehensible to me. ‘Every individual sees color differently’, yes, but this didn’t prevent Messenger’s team to produce true color movie of Earth; this does not prevent the production of true color pictures of Mars, Jupiter, Saturn or of flowers in the field after all.
2. The processing of channels, as described in the caption, looks quite arbitrary and unlinked with the reproduction of true color, perceived by human eye. Meantime, old Mariner-era color photos of Mercury, although based on only two wavelengths, were processed to be true-color, and are brownish, more similar to upper right image in PIA11364.
3. I am a full profane in image processing, but there is ‘ugordan's photostream’ of Gordan Ugarkovic, who speaks about calibration, spectral characteristics, etc. His shots of Mercury are brownish too.
I admit readily that I may be completely wrong, but what if I’m not?
Bye!
P.S. Strangely, some of previous of Steve Albers' Mercury textures had a patch at right, corresponding to flyby 1 outbound shots, of much better resolution.
P.P.S. What do you think about ugordan's picture of true-color Venus?

[jogad]

Hello,

I like textures better with the unknown regions filled in instead of having blank areas.

You should know how very much I prefer the realistic approach by now.

The miracle with Celestia is that there is no reason to make a choice.

When I asked Vorthon in another post to make a texture to fill the gaps in Mercury I was only half kidding. But my words were misinterpreted and I did not have a response from him.

So I made myself the texture I wanted from the one that Fridger gave to Celestia.

It is absolutely not a texture to replace Fridger's one. The purpose and the audience are different.
The "normal" map answers the question "What the planet looks like?" and aims for example at people who want to discover astronomy. This is a good tool for vulgarization.
On the other hand, the “Limit-of-knowlegde” map is very useful to knowledgeable people to find what parts are already explored and to show others that human knowledge is an ongoing process.

For example, to briefly introduce the planets of the Solar System, it is unnecessary to know which parts of the surface are actually known or not. In this case, my view is that it is better to show a planet as alike as possible but witch is not cause of unwanted questions. If necessary, it is always possible to show the really known world map.
I think that both maps can be useful depending on the circumstances.

Saying that a map is an l-o-k map is not an insult. L-o-k maps represent the beauty of the hidden truth behind appearances.
But often they are appearances that make people want to know what is behind.

Here is the result. There are mainly two unknown areas that I filled by copying known parts of the same texture.

Since I am not an artist, I would be happy if these pictures could inspire someone to do a better job.

First main region in the northern hemisphere.



Second main region in the southern hemisphere.



regards

[scalbers]

Looks like Fridger's revamped version should look pretty good in Celestia - particularly with those white blank areas and other seams taken care of. It's hard for me to tell - though perhaps some subtle sharpening might be helping here as well on the lower resolution imagery. I agree also (with RVS) that it's worth considering the pros and cons of Ugordan's recent color versions of Messenger flybys.

Steve