High Resolution Texture Of Venus

[scalbers]

Fridger,

I note that some of the other higher resolution images mentioned in this thread also have some artifacts present. I reviewed my software a bit to see if there are any easy ways for me to suppress them. Many of the artifacts are discontinuities between the Magellan data and the underlay (from Calvin Hamilton) that includes other spacecraft. I had adjusted the Magellan overlay based on average brightness and contrast over the entire map. Naturally this broad-brush approach will not allow the brightness/contrast fit to be as good as it could be. Perhaps having a regionally varying brightness/contrast of the Magellan overlay will help it to blend in more smoothly.

Steve,

that matches about my own investigations of the matter. But I am surprised why you did not merge the various layers from different sources using standard soft-mask techniques which eliminates such artifacts right from the start.

I think there are two complementary strategies as to textures, depending strongly on their purpose:

--planetary scientists would want to leave such transitions exactly unmodified in order to retain the complete information an image contains. They rather don't seem to care too much about such nasty seams.

--applications like Celestia want to first of all create a breathtaking yet realistic 3d simulation of celestial bodies. In this case nasty seams, stripes and all that are highly /undesirable/.

Bye Fridger

Fridger,

Glad to see this discussion continue. Andrea's map looks pretty good to me as well. I'd like to eventually compare a 4K (perhaps TIFF) version to mine to get a feel for the processing that was applied.

My software is written pretty much from scratch in IDL, so I'd have to learn a little bit more about standard soft masking techniques and the best way to code them. Any quick online references to them that you are aware of?

Perhaps I'm trying to straddle the line between the planetary science and the more artistic community - I may run the risk of pleasing neither as a result ??

[ANDREA]

No CRC error.But unfourtanetelly all textures are black and white,and for some reason,there is no way to colorize it under Photoshop.Strange,because I had got it before.Can someone help me or give where I can find a color 8k or 16 K Venus texture?

Danjeli, just now I downloaded and opened without any problem the JohnVanVliet's 4k.png and 8k.png Venus textures (both very nice and detailed, BTW!) from MotherLode Creators page.
Moreover both are colored, in a color similar to my 4k.DDs texture's, but softer in contrast (and John's is rotated 180 deg, as requested by Fridger).
I think that John's 16k texture will give same positive results.
What kind of settings has your PhotoShop?
Why don't you try again with PS?
Bye

Andrea

[Malenfant]

As Fridger correctly says, sometimes it's necessary to add something to the known realty, so I made a long work on the radar map in order to eliminate all the artifacts, defects, missing areas in the original.

That's funny because usually Fridger is a proponent of realism, not aesthetics. I seem to recall him getting into arguments about whether to use Hubble's blobby Pluto texture or a fictional one. Or in having Limit of Knowledge masks. Or complaining when people preferred anything less than his realistic binary stars and galaxies.

And yet now he's OK with a 'fake' Venus with extra fabricated cloned bits covering over the limits of knowledge to make it look pretty? I think he must have been replaced by an imposter!

[Malenfant]

My software is written pretty much from scratch in IDL, so I'd have to learn a little bit more about standard soft masking techniques and the best way to code them. Any quick online references to them that you are aware of?

Perhaps I'm trying to straddle the line between the planetary science and the more artistic community - I may run the risk of pleasing neither as a result ??

I think the obvious solution is to have two textures - one that is realistic and reflects the limit of knowledge (complete with image artifacts), and one that is "pretty", that hides the artifacts but creates fictional data in the process (even if it's just making different resolutions look like they're the same, or sharpening images).

You using ISIS, BTW?

[t00fri]

As Fridger correctly says, sometimes it's necessary to add something to the known realty, so I made a long work on the radar map in order to eliminate all the artifacts, defects, missing areas in the original.

That's funny because usually Fridger is a proponent of realism, not aesthetics. I seem to recall him getting into arguments about whether to use Hubble's blobby Pluto texture or a fictional one. Or in having Limit of Knowledge masks. Or complaining when people preferred anything less than his realistic binary stars and galaxies.

And yet now he's OK with a 'fake' Venus with extra fabricated cloned bits covering over the limits of knowledge to make it look pretty? I think he must have been replaced by an imposter!

No, I am fine with this approach more or less. But Venus is really special. A bit related to the texture mapping of Titan.
Also here the total cloud cover in visual light and the surface mapped in a near IR wavelength window.

Analogously:

1) the /realistic/ approach for Venus is just the untransparent cloud cover. Fine.

2) The "unrealistic" approach is a radar scan that just uses another wavelength than visual light. It traverses the clouds and maps the surface differently. The little extra stuff here and there ...hmm --too bad.

Bye Fridger

[Malenfant]

Analogously:

1) the /realistic/ approach for Venus is just the untransparent cloud cover. Fine.

2) The "unrealistic" approach is a radar scan that just uses another wavelength than visual light. It traverses the clouds and maps the surface differently. The little extra stuff here and there ...hmm --too bad.

Bye Fridger

But why comprimise on the surface texture though, just because it's not technically visible in visible light? We have real data for what's under the clouds after all - isn't that worth displaying realistically too?

(From a realism perspective, I do hate the orange colour of the Venus surface textures too. Particularly since it has no colour in radar imaging (keep it greyscale!), and in visible light the surface doesn't even remotely look like that - it'd be brown/black like most normal basalts are)

[t00fri]

Analogously:

1) the /realistic/ approach for Venus is just the untransparent cloud cover. Fine.

2) The "unrealistic" approach is a radar scan that just uses another wavelength than visual light. It traverses the clouds and maps the surface differently. The little extra stuff here and there ...hmm --too bad.

Bye Fridger

But why comprimise on the surface texture though, just because it's not technically visible in visible light? We have real data for what's under the clouds after all - isn't that worth displaying realistically too?

(From a realism perspective, I do hate the orange colour of the Venus surface textures too. Particularly since it has no colour in radar imaging (keep it greyscale!), and in visible light the surface doesn't even remotely look like that - it'd be brown/black like most normal basalts are)

I have always advocated a multi-wavelength display mode of Celestia. So I am certainly happy with a radar scan. But radar data have often little to do with the actual morphology of the surface. Bright and dark areas mean something different as we all know...

Also I am not so sure whether Andrea's coloration is so unrealistic, since the surface color MAINLY reflects the color of the diffuse light filtered by the dense cloud cover (cf Titan! <-> orange) . So the clouds being yellowish, may well produce a warm (yelow-brownish) coloration of the gray rock on the surface.

Certainly for me Andrea's color is as "good" as grayscale, both being not right

But independent of this color issue, I just dislike such artefacts that may easily be avoided in the assembling process of the full texture.

Personally, I never invested much time in working on Venus textures, precisely, since I am not yet certain how to do a "realistic" representation, that may be acceptable from a more scientific point of view...

Bye Fridger

[Malenfant]

Also I am not so sure whether Andrea's coloration is so unrealistic, since the surface color MAINLY reflects the color of the diffuse light filtered by the dense cloud cover (cf Titan! <-> orange) . So the clouds being yellowish, may well produce a warm (yelow-brownish) coloration of the gray rock on the surface.

Yes, but if the surface is dark grey/black/dark brown (likely if it is made predominantly of basalts) then it's still going to look that colour regardless of the diffuse light seeping through the clouds. A black rock in a diffuse yellow light still looks black!

Plus, this is rather moot for radar since the colour of visible light reflected by the surface and the colour of the diffuse light coming through the clouds doesn't affect the radar response at all. The bright surfaces on the radar map are actually either rough surfaces or regions with minerals that have higher emissivity - the dark surfaces on radar are actually smoother surfaces or regions with lower emissivity. While the pixel intensity/greyscale value of the pixels on the map reflect the radar properties of the surface, colour value has no meaning in the radar map - it's an entirely artificial addition done purely for aesthetics.

But independent of this color issue, I just dislike such artefacts that may easily be avoided in the assembling process of the full texture.

I guess it depends what one's goals are. In a scientifically realistic texture I'd avoid any cloning of the texture or addition of material that wasn't there - I'd just use the magellan images and use lower resolution radar data to fill the gaps.

Personally, I never invested much time in working on Venus textures, precisely, since I am not yet certain how to do a "realistic" representation, that may be acceptable from a more scientific point of view...

Ideally a realistic texture would be made by somehow converting the radar data into their visible equivalent, but that's impossible right now. But since radar data is all we have I think that should serve as a 'realistic' texture, since it's what we know and what we can get.

[ANDREA]

Yes, but if the surface is dark grey/black/dark brown (likely if it is made predominantly of basalts) then it's still going to look that colour regardless of the diffuse light seeping through the clouds. A black rock in a diffuse yellow light still looks black!

OK, Malenfant, trying to make a bit more science about the color of Venus surface, whose we have some (ugly!) B&W images from Venera 9 and following Veneras, and this horrible color image from Venera 13:



there are many doubts on the effective color, because Venus atmosphere cuts blue light.
But speaking of basalt rocks (and, for similitude, of igneous rocks), here are three specimens of Earth basalt rocks:



The image of my Venus texture is very close to the right rock:



while this one is close to the middle one:



and this one is very close to the left one:



What is the final result, IMHO?
Up to now noone knows how Venus basaltic rocks appear, after having been subjected to an infernal acid rain-winds-heat-pressure environment.
We can only suppose that the color is similar to the Earth basaltic rocks, and here we are back to my three examples.
But, sorry, I remain of the opinion that the first image is more "natural" than the others.
IMHO, naturally.
Bye

Andrea

[t00fri]

Enlightening!

Cheers,
Fridger

[jestr]

Hi Andrea,dont know if you have seen this web page,but it has some slightly different versions of the Venera13 and 14 missions
http://www.mentallandscape.com/V_DigitalImages.htm
cheers,Jestr

[Malenfant]

What is the final result, IMHO?
Up to now noone knows how Venus basaltic rocks appear, after having been subjected to an infernal acid rain-winds-heat-pressure environment.

The Venera photos show what the rocks look like - they look like ordinary basaltic rocks.

We can only suppose that the color is similar to the Earth basaltic rocks, and here we are back to my three examples.
But, sorry, I remain of the opinion that the first image is more "natural" than the others.

But that's the thing - all you're doing is tinting a greyscale radar map so that the dark parts are similar in colour to a basalt rock. But this is unrealistic because you're mixing incompatible data. Radar doesn't show what colour a rock is, it just shows how smooth/emissive it is.

If the atmosphere of Venus was to suddenly disappear so we could see in visible light the surface, we wouldn't see the bright rifts and arcs and fault zones and mountains and highlands that we see in the radar map - all those regions would be dark in visible light because they're still just made up of basalt and volcanic rock which is dark.

But as yet we can't see through to the surface of Venus in visible light - all we have are the Venera pictures which are looking at a dark surface through an atmosphere that adds a yellowish tint - but the colour of the surface material itself remains dark grey/brown/black!

So do you see what I'm trying to say here? You're adding a colour to a radar map that has no right to be there- it's incompatible with the radar data that we are using to make the texture.

The only reason NASA added that orange colour to the radar map was for aesthetics - it in no way reflects the real colour of the surface itself. For both Venus and Titan, even if you had visible light data from the surface, any colour that you add would be colour that is added by light coming through the atmosphere, not of the surface itself. And I don't think that colour added by atmosphere should be added to a texture map.

[ajtribick]

Presumably there are topographic maps of Venus available, so one possibility might be using the topographic map as a bump/normal map, though that would have the problem of only being visible near the terminator.

Perhaps some fairly low contrast version (like ANDREA's third one perhaps), plus a bumpmap might be an acceptable compromise?

[Malenfant]

Presumably there are topographic maps of Venus available, so one possibility might be using the topographic map as a bump/normal map, though that would have the problem of only being visible near the terminator.

Perhaps some fairly low contrast version (like ANDREA's third one perhaps), plus a bumpmap might be an acceptable compromise?

There is altimetry data available as greyscale images (white = high, black = low) on the Magellan CDs.

There's a gif available of it here, but it needs to be resized for Celestia. However the data gaps in it were filled in with fictional data cloned from nearby.

[t00fri]

The only reason NASA added that orange colour to the radar map was for aesthetics - it in no way reflects the real colour of the surface itself.

Right, of course. That reddish brown "radar" color suggests HEAT HEAT and again HEAT on the surface

For both Venus and Titan, even if you had visible light data from the surface, any colour that you add would be colour that is added by light coming through the atmosphere, not of the surface itself. And I don't think that colour added by atmosphere should be added to a texture map.

Right again: no atmospheric effects into textures! In principle...Until Chris finally finds the time to redo COMPLETELY the present atmosphere code! It's high up on his todo-list, but who knows when there will be respective action...

Bye Fridger

[t00fri]

Hi Andrea,dont know if you have seen this web page,but it has some slightly different versions of the Venera13 and 14 missions
http://www.mentallandscape.com/V_DigitalImages.htm
cheers,Jestr

Hi Jestr,

that's a pretty valuable site...

Thanks for the pointer.

cheers,
Fridger

[danielj]

Nope.I downloaded it and it was in black and white.I downloaded form index of creators:
http://celestiamotherlode.net/creators/johnvanvliet/
I downloaded the 16kNewVenus.zip file.

No CRC error.But unfourtanetelly all textures are black and white,and for some reason,there is no way to colorize it under Photoshop.Strange,because I had got it before.Can someone help me or give where I can find a color 8k or 16 K Venus texture?

Danjeli, just now I downloaded and opened without any problem the JohnVanVliet's 4k.png and 8k.png Venus textures (both very nice and detailed, BTW!) from MotherLode Creators page.
Moreover both are colored, in a color similar to my 4k.DDs texture's, but softer in contrast (and John's is rotated 180 deg, as requested by Fridger).
I think that John's 16k texture will give same positive results.
What kind of settings has your PhotoShop?
Why don't you try again with PS?
Bye

Andrea

[ANDREA]

Hi Andrea,dont know if you have seen this web page,but it has some slightly different versions of the Venera13 and 14 missions
http://www.mentallandscape.com/V_DigitalImages.htm
cheers,Jestr

Hello Jestr, yes, I have seen that page, thank you anyhow, but I preferred to use the raw original image, because the author here says "with a little reprocessing".
If you compare it with the original image as issued by Soviets

Code: Select all

http://nssdc.gsfc.nasa.gov/imgcat/html/object_page/v13_vg261_262.html

you can see it's very different, with a lot of greens that were invisible in the raw one.
But here we are back at the starting question:
what color is real?
The color seen with unaided eyes?
Or the color seen through a series of standard filters (what" standard")?
And if you have neither possibility, but have an idea of the rocks type, the same color of the same rocks on the Earth, without considering the different local environment?
Let me give you an example: we have zillions of wonderful Mars surface images, but we are still discussing on Mars colors.
Many years ago I was an active plastic modeller, and one of the most important things to consider before painting an aircraft was that the given FS colors were for an aircraft seen closely, not far away, so that a scale 1/100 model needed darker colors than a bigger 1/32 scale one (because apparently closer).
Many times we have no real possibility to give a correct interpretation, so we try to make our best, along with our knowledge.
BTW, when I showed the usual radar surface of Venus with the original red-orange-yellow colors, many times people asked me ".. are those active vulcans spreading lava flows?"
This is the reason why I modified the radar map to the new one.
Sorry if this is not science, but it's the closer thing we can do on this subject, IMHO.
Bye

Andrea

[ANDREA]

Nope.I downloaded it and it was in black and white.I downloaded form index of creators:
http://celestiamotherlode.net/creators/johnvanvliet/
I downloaded the 16kNewVenus.zip file.

I downloaded today all three textures from the same page, and 4k and 8k are in color, and both cleaned by John, as you can see here
4K

8K


The 16kNewVenus.zip file is in B&W (but I had no problem to open it), but it's a crude texture, with all the artifacts, missing parts, errors and so on.
If you cannot see the 4k and/or 8k in color, sorry for you, Danielj, but surely you have some problem in your PC, or PhotoShop, or something else.
I suggest you to check if in the Photoshop settings there is something that doesn't allow .png file in color, or if you have "Image" "Mode" "Grayscale" enabled.
Good luck.

Andrea

[Malenfant]

But here we are back at the starting question:
what color is real?
The color seen with unaided eyes?
Or the color seen through a series of standard filters (what" standard")?
And if you have neither possibility, but have an idea of the rocks type, the same color of the same rocks on the Earth, without considering the different local environment?

These are good questions. I know I've looked at many images from spacecraft of planets and moons and thought "we'd probably never see these moons like this with our eyes - these are images taken over long exposures".

But in the case of Venus (and Titan), we know what ISN"T real colour. What we're seeing there is radar reflectivity and surface smoothness. Adding colour to it makes no sense whatsoever (unless it's actually as an extra dimension to the image - say, as colour-coded topography overlaid on the radar image).

The local environment is irrelevant - it doesn't magically change the rocks. They're mostly basalts just like on Earth. Maybe they have slightly different chemical details, but they're basalts nonetheless. Even at the temperature and pressure on Venus' surface there's nothing in them that would be radically different, and we know the properties of a hell of a lot of minerals and how they change with temperature very well. So the colours aren't going to be much different, and the Venera surface images don't seem to imply that they are either.

Let me give you an example: we have zillions of wonderful Mars surface images, but we are still discussing on Mars colors.

Yes, because we have a range of different visible light filters through which to look at Mars (both on the surface and in orbit, and also from Earth itself). But we don't look at Venus like that - we've got global datasets that are radar, and we 've got a few photos from the surface. For Venus we have to extrapolate a bit - we know that the surface is made of volcanic rock - mostly basalts, which are dark in colour. There are undoubtedly other types of rock (again, all volcanic in origin) maybe of different colours - some brighter, some darker - like on any other planet. But we know for certain that the brightness of the terrain we see in the radar images do NOT reflect the true colour or brightness of the surface in visible light.

BTW, when I showed the usual radar surface of Venus with the original red-orange-yellow colors, many times people asked me ".. are those active vulcans spreading lava flows?"

There might well be active volcanoes on Venus stiill. Just not for the reasons that the people assume .

This is the reason why I modified the radar map to the new one.
Sorry if this is not science, but it's the closer thing we can do on this subject, IMHO.

Well from an aesthetic viewpoint I do prefer your map colours, they're less garish than the bright orange/yellow of the usual maps. And certainly any 'visible light map' of Venus is going to be largely speculative. But I think it's still important to have these discussions so that everyone's on the same level of understanding about what we know