Celestia Forums

Well it has been mentioned to me twice now by thoughs that have seen the Mars image called Valmar that the volcanoes on the Tharsis platue look dark. Well after going over several pictures from NASA I have come to only one conclusion. What we are seeing in our Mars textures is all wrong.
The volcanoes are in fact much darker than the surrounding area. To push the point home here are some pictures from a NASA website about weather on Mars. Have a look at these.







If these don’t convince you than I don’t now what will.
So as always I decided it was time to fix this error. I used Jim's new blending of his low shadow texture and Space-Graphics new shadowless texture as the base.
After a few hours of work here is the result. I was quite shocked at how good this turned out. I feel this is probably the closest texture to actually being there. Here are some sample shots.




I have just a little tweaking left and then I will make it available in several versions and sizes.
Don.

Hi Don,

Your modification looks very interesting.

Don. Edwards wrote:Well it has been mentioned to me twice now by thoughs that have seen the Mars image called Valmar that the volcanoes on the Tharsis platue look dark. Well after going over several pictures from NASA I have come to only one conclusion. What we are seeing in our Mars textures is all wrong.

I searched on Nasa's Planetary Photojournal and found most of your Mars pictures. I noticed that most of the latest colored Mars pics has now this dirty brown. But have Mars really those colors?
I found some interesting statments in the description of the pictures.

http://photojournal.jpl.nasa.gov/catalog/PIA03229

"The Image above is one of a series of simulated views of Mars as it would be seen from the Mars Global Surveyor space craft."

http://photojournal.jpl.nasa.gov/catalog/PIA03755

"Mars Global Surveyor (MGS) orbits around the red planet 12 times a day. Each orbit goes from pole to pole. Over the course of a single day, the wide angle cameras of the Mars Orbiter Camera (MOC) system take 24 pictures--12 red and 12 blue--that are assembled to create a daily global map."

Aha, most of the latest pictures are made by Mars Global Surveyor and built from a red and blue channel. The sythesis of the green channel was very simple only an the average of red and blue.

Now I have some doubts about the dirty brown looking of Mars.

Here is another picture:



"Hubble can see details as small as 10 miles (16 km) across. The colors have been carefully balanced to give a realistic view of Mars' hues as they might appear through a telescope."

Now you can see the answer about the colors of Mars is not simple and may depend on sem personal ideas.

Bye Jens

jim wrote:Hi Don,

Your modification looks very interesting.

Don. Edwards wrote:Well it has been mentioned to me twice now by thoughs that have seen the Mars image called Valmar that the volcanoes on the Tharsis platue look dark. Well after going over several pictures from NASA I have come to only one conclusion. What we are seeing in our Mars textures is all wrong.

I searched on Nasa's Planetary Photojournal and found most of your Mars pictures. I noticed that most of the latest colored Mars pics has now this dirty brown. But have Mars really those colors?
I found some interesting statments in the description of the pictures.

http://photojournal.jpl.nasa.gov/catalog/PIA03229

"The Image above is one of a series of simulated views of Mars as it would be seen from the Mars Global Surveyor space craft."

http://photojournal.jpl.nasa.gov/catalog/PIA03755

"Mars Global Surveyor (MGS) orbits around the red planet 12 times a day. Each orbit goes from pole to pole. Over the course of a single day, the wide angle cameras of the Mars Orbiter Camera (MOC) system take 24 pictures--12 red and 12 blue--that are assembled to create a daily global map."

Aha, most of the latest pictures are made by Mars Global Surveyor and built from a red and blue channel. The sythesis of the green channel was very simple only an the average of red and blue.

Now I have some doubts about the dirty brown looking of Mars.

Here is another picture:



"Hubble can see details as small as 10 miles (16 km) across. The colors have been carefully balanced to give a realistic view of Mars' hues as they might appear through a telescope."

Now you can see the answer about the colors of Mars is not simple and may depend on sem personal ideas.

Bye Jens

I think Jens addressed an important point which was on my mind already since a while! Being an active amateur astronomer since my early youth, I have looked at Mars many many times using telescopes of a very large range of diameters ( up to > 1meter!). Mars always looked orange to /red/ to me. Certainly not like Don's yellow dirt color! Much more like the Hubble picture above.

Bye Fridger

Hey Guys I simply desaturated the texture a little. I seem to be having a problem with my vertex shaders as they seem to changing the color and inverting the bumpmaps. I have to keep cycling through the vertex options till I get something that looks ok. I am not really very happy with the final 1.30 version of Celestia. For some reason it won’t keep my chosen setting for the vertex shaders. I took a screen capture movie of the texture and the color in it was right on the mark. Besides this was about the volcanoes not my textures color. As I said I had some tweaking to do. Like put the color back to normal. It was easier for me to do the work I was doing with the color cooled down a little.

Ok here the color corected version. As you can see the color is right back were it started.


And this is with jim's Mars clouds.


The Clouds now really play off the volcanoes much more like what is seen in many of the about photos. One more thing on the volcanoes beeing dark and not light in color. There are even black and white images from the Marineer and Viking probes that show the volcanoes as dark. If I can locate them I will post them.

Don,

The vertex shader code in 1.3.0 final does have problems. I'd suggest using 1.3.0pre6 until 1.3.1 is available.

If I didn't delete it. I was just about to plug the texture into 1.2.x and see how it looked in there. Thanks for the info.
Don

It's still available in
http://www.shatters.net/celestia/files/ if you don't mind downloading it again.

Is there any color difference when viewing mars through earths atmosphere as opposed to from space?

Ok,
Last picture. Now that I am using am older build things are a little better.
Here is the last sceenshot from me.

Now everyone else can play with it and see what they think.
So here are the textures:

2k in PNG only with bumpmap
http://63.224.48.65/~impulse/Mars_DrkMn ... Mnts2k.zip
4k in DDS and PNG with bumpmap
http://63.224.48.65/~impulse/Mars_DrkMn ... Mnts4k.zip
8k in DDS only with bumpmaps in 2 sizes.
http://63.224.48.65/~impulse/Mars_DrkMn ... Mnts8k.zip
Any feedback is always welcome.

Don.

Well here is a slight explanation as to why my screencaps on my textures changes so much. With the changes Chris made to the way Celestia handles Vertex shading and the fact he took the shortcuts out of the menu, every time I start Celestia I now have to ctrl-V to get the vertex setting I want. I myself liked the old way better. Its a royal pain having to do this everytime I start the program. So I am going back to an even older build that is more in tune with my NVidia card. At least in the old builds I didn't have to keep going into the menu to turn the vertex shaders on. For some reason the newer builds simply will nopt hold my chosen settings. That is why often my posted pictures are shifting around in color. To prove the point have a look at the folowing pictures.

Basic Mode


Multitexture Mode


OpenGL Vertex Mode


NVidia Combiners Mode


OpenGL Vertex/NVidia Vertex Mix Mode


As you can see most don't show a great deal of change but when you look at the OpenGL/ NVidia mix mode there is quite a shift in brightness and color. Again this is one of the reasons why my textures looked off but not the only reason. As I said I am going back and doing some testing in a much older build. I don't know if its Celestia or the drivers I installed recently but there is definetly something going on here that I never had a problem with in the past. There is also one other posibility. I recently changed monitors to a giant HP 21" Trinitron and its color has needed to be rather agresinvly changed. Again this could be an issue with color but not for my chosen vertex setting not being held. So until I figure this out I don't want to post any furtheer textures till I have this nailed down. But my main focus for now is this monitor. I will be hooking up my good old 17' Trintron superflat and seeing how things look with that. Wish me luck.
Don.

Don,

Unfortunately, as I suspect you know, most CRT manufacturers set monitors and TVs to be very blue by default. (i.e. a color temperature of 9000 or so) It makes them bright and eye catching, but not very accurate at all. These settings may be OK for text and 3d layout, but they're a disaster if you want accurate colors.

Some companies include adjustments so you can set the color temperature to be a more appropriate lower value (~6800?). Sometimes you have to use a calibration DVD like "Avia Guide to Home Theater" to get the RGB settings right. Much more expensive auto-calibration hardware is also available that uses a sensor to monitor the colors generated by the screen. *sigh*

Also, is there any chance you could edit your posts to use thumbnails that link to the larger pictures? Having to wait forever while the pictures download and watching the browser keep changing the location of the text gets to be rather, well, annoying. I think most of the color effects don't need such large images after they've been seen once.

Thanks!

selden,
My monitors native color range fortunately isn't that blue. That was my error. I used NVidia's color tools to force a more crisp bluer image and this monitor really doesn’t look good with those settings. I set everything back to standard and this seems to help a great deal. One thing I wish microsucks could do is at least developed a way to calibrate your monitor like Apple does with ColorSync. This would have helped from the beginning.
Unfortunately its just to much of a hassle to work on the textures on my PowerMac and then transfer everything over for testing back and forth. And unfortunately my PowerMac is just to old and graphically challenged to run Celestia on under OS X.
I will try to remember about using smaller images. I think there are quite a few of us that are guilty of this particular crime.

Hi i used this nasa photo to creat a color gradent to colorise a grayscal map
the final image (after somt color correcting ) was this map

I will see what i can do about darkning the valcanos

Oh John,
You shouldn't have posted that. Now Fridger is just going to make comments about your choices of colors. He and I have been back and forth over a number of issues. He will not comment on the "Dark Volcano" issue. But that’s fine as I have someone checking this with NASA and that issue will be put to rest once and for all. He will no doubt bring up that these pictures were taken in only the red and blue color range and that the green channel had to artificially be made and etcetera and so on………but that sure doesn’t explain the drastic contrast differences between the volcanoes and the surrounding Tharsis rise. I am sure he will say that it is just an optical illusion or something. As I said we will soon see if I am right or wrong on this. If I am wrong so be it and I will have to rework my texture.
He sure has been willing to take issue with my choice of texture coloring. I have of course have been taking issue with his texture colors as well. I made a few compromises that I feel are a better choice although I had to bite my lip on a few things. I still feel his colors are just a tad to dark. His colors do not allow for a proper albedo effect of Mars from a distance. Mine do however. And while he keeps comparing the screen captures of his texture to Hubble photos if you look close you can see that the texture color is darker than in the photos. He also isn't taking into consideration that there are different areas on Mars were the topography is different colors. He has basically made all Mars a very dark rusty red with no allowances for these differences. The volcanoes are one area to point at as well as the Tharsis rise witch tends to be brighter, some of the southern highlands, and the Hellas basin is always brighter as well. Of course I can’t make a final judgment until he has it up for download and I try it myself when ever that will be. I know he is limited to a modem so uploads for him are difficult unlike those of us with a broadband connection. As in all things it will have to be a wait and see thing.
But by all means keep working on your textures as I keep working on mine and Fridger on his. Somewhere there in the middle is what Mars really looks like. I wish we could get some very nice 35mm pictures of Mars but I am afraid that is just not possible. So we have to rely on pictures from NASA and the Hubble and who’s to say if any of them are right anyway.

Don.

Hi I tend to go for visualy pleasing and not neceraly 100% correct
The light valcanos are from using the bump map for the details
I am corecting this with a 16k gray map from PDS map a plenet