New Mars Forums

I've read some people have doubt over the calim of 30cm res - based on the limiting resolution of the optics - anyone got any ideas on that?

Doug

Hadnt thought of posters.  Good call.

All in good time.

Doug

Never seen it ( never looked for it to be honest )

Doug

I think thats an APXS impression - it's made several before.

APXS works only over a TINY distance of a few mm ( it doesnt work at all in an Earth atmosphere - as that's too dense )

Doug

Well - I'm no scientist either - I'm a tempremental artist

The Raw jpg's that jpl puts out at the moment are useless. One can make NO claims about sky colour or any other colour for that matter. - I'll do my best to explain why.

Each image is treated individually to what ammounts to an automatic adjustment to show the best range of brightness.  That means that even if the brightest thing in one image is infact only a mid grey - it'll be stretched so that it's white and everything stretches underneath it.  In photoshop this is called 'auto levels'

Unfortunately - they do it to each image of each filter - ignoring its context to its matching images thru other filters.

Thus - if you're looking at nothing but sky - an L2 image ( near infrred) will be an averaged grey noise.   So will L4, L5, L6 and L7.

Merge that in photoshop to get colour - and you get nothing but grey - because they're all averaged out the same way - even though in actual fact - the red channel should be brighter than the blue channel say.

Now - part of that auto-levels treatment is done on the rover itself. It takes greyscale images with 12 bit depth  ( 4096 shades of grey ) -  using exposure either specifically commanded, or its best guess based on previous images.  It often then does a bit of a stretch on the image to turn it into an 8 bit image that is obviously, smaller in file size.

That then gets sent down to the ground - and reverse engineerd based on what the rover did to it to get a proper image that is an accurate representation of the brightness with that given filter.  When you merge these in photoshop - you get a fair representation of the colour on mars.

However - to give the best dynamic range - they auto-level them agian here so at least one pixel is white and one pixel is black and the rest is stretched in between.  THAT - is what you get with the Raw JPG's.

to give you an idea - this is an image made using L4, 5 and 6 filters of the Columbia dedication on the back of the HGA ( I know - a morbid target to test the issue - but an excellent one technically - with lots of variation in colour )

This is made using the raw JPG's from the nass website

This is made from calibrated IMG files from the PDS

See how the former has at least one pixel that is totally white - because of that auto-level treatment they've had.  The properly calibrated file however - has no such tendency and looks far more 'natural'

This hue also matches that at the bottom image of these choice pieces of work by Justin Maki ( part of the MER team now ) of the Pathfinder field of view http://mars.jpl.nasa.gov/MPF/science/PD … jun99.html  - notice the picture of the airbages look just like the one I did of the back of the HGA on Spirit. Notice the sky colour.  These are part of a peer reviewed paper on the subject.

Another comparisont that shows JPL raw JPG's - and proper PDS IMG's

See how things change from frame to frame from the JPGs. Notice also that because there is actualyl somethign white in the bottom frame - the auto-levels did little to ruin the bottom frame. You will find that images including airbags, sundial or IDD will usualyl come out well because something is actually white in the field of view.

Also note how the colour between the top and bottom frame is identical in the left set. These have been properly calibrated and released according to PDS rules and peer review at JPL - and it is from images like this that one can begin to make a judgement on the colour of the martian sky.

No - I dont think the sky would be bright red.  I think its salmon pink. One thing is certain - the sky is packed full of dust. What colour is the dust? Salmon Pink / Beige.  What colour's the sky therefore - Salmon Pink / Beige.

The closer to the sun you look however - the less you're seing dust that's lit by the sun, and the more you're seing light thats pasing THRU the dust instead - refracted if you will- and this can cause it to look a little more like it would if there were no dust at all - a little more grey - bordering on blue.

There is no conspiracy to make mars look more alien. There are some excellent papers that discuss the issue of proper photometric studies of mars.

For the very simplist test - look at the sundial on the rovers.   If the sun dial is looking about right in terms of colour - then look at the two little mirrors - one each side just outside the white band

THAT - is the colour of the sky on mars - and when most of the sky is that colour - then the whole place is going to look that sort of colour (hint - what colour does a white piece of paper look on the floor of a room with bright blue walls - slightly blue ! )

This guy explains it more scientifically and quantitatively than I could ever dream of doing

http://www.atsnn.com/story/30048.html]h … 30048.html

So - conclusions.

1) Dont pretend you can make a call on colours based on the raw JPG's. You cant. They're not a scientific product.

2) Nasa isnt doctoring colours. Pancam team alone runs into dozens of people. Consider the Viking people, the Pathfinder people, and the MER people - there's THOUSANDS of people been involved in the seemingly simple task of taking photographs from the surface of mars. Are we to believe that they're doing the best job that science can of showing us what colour the place is - or are we to believe that actually - they're ALL part of one massive conspiracy to make mars a different colour than it really is.

3) The sky is salmon pink / beige typically. It varies a LOT however - depending on where you look ( with respect to the sun ) and how dusty the atmosphere is (huge seasonal variations depending on dust storms) - and where you are ( Meridiani and Gusev look different to one another, and different over time as the volume of dust in the atmosphere changes)

So there we go - no sexing up of mars is going on   And when I make my pictures from the raw jpg's - all I ever do is adjust them to  look as 'nice' as possible. Not accurate, not right, not fudged. Just ''nice'.

Doug

Hi guys - someone pointed me this way - I'm the chap who starter the mer.rlproject.com forum and who's pics you've been enjoying

Quick note on colours - Basically, I do the best I can based on the rubbish JPG's that JPL put out as 'raw' data. They're usually OK when you have a lot of white in the scene - say looking at the lander airbags - but when there's no simple white point for the auto-gain to lock on to, then the raw imagery varies terribly and the closer you get to the sky the bluer things become.   The proper raw data has begun to be published, and I intend to do a series of some small panoramas based on it as and when they appear -  probably no 360 degree pieces, they're just too big for my machine ( 24,000 x 4,000 anyone) - but certainly  anything from one frame to 20 or more My grand plan is to write a book in the same mould as 'Full Moon' by Mike light.

Anyhoo - not seen it mentioned - you've seen the SW view from Clovis ( the line of three or four frames looking out over the rock onto the floor of gusev ) and the view to the NE - (the larger mosaic looking along the edge of the Columbia hills )

HERE - http://mer.rlproject.com/index.php?show … wtopic=334 is the view looking roughly NW back toward Bonne.

Incase some of you havnt trawled back thru the old pages on the forum - here's a few of my fav works to date

Oppy :
Burns Cliff - http://mer.rlproject.com/index.php?show … wtopic=201

Super Res of Backshell - http://mer.rlproject.com/index.php?show … wtopic=175

First Trench - http://mer.rlproject.com/index.php?show … owtopic=77

Rat + Hole - http://mer.rlproject.com/index.php?show … owtopic=67

Early Outcrop - http://mer.rlproject.com/index.php?show … owtopic=23

Spirit ( by far the more talented photographer )

Columbia Approach -http://mer.rlproject.com/index.php?showtopic=247

Drive and Turn - http://mer.rlproject.com/index.php?show … wtopic=238

Dont look back - http://mer.rlproject.com/index.php?show … wtopic=202  ( still my desktop image - various sizes here - http://mer.rlproject.com/index.php?show … wtopic=203 )

Distant Tracks - http://mer.rlproject.com/index.php?show … wtopic=193

MONSTER TRUCK - http://mer.rlproject.com/index.php?show … wtopic=181

Super Res imaging of Heatshield - http://mer.rlproject.com/index.php?show … wtopic=177

Hows about photoshop? http://mer.rlproject.com/index.php?show … wtopic=117

Empty Nest - http://mer.rlproject.com/index.php?show … owtopic=38

Anaglyph that works -http://mer.rlproject.com/index.php?showtopic=32

Enough spamming from me - any q's about the piccies - let me know

Doug