Exploring Gediz Vallis and its ridge, Sols 3871-xxxx, 1 Jul 2023- Options

[jvandriel]

The Navcam R view on Sol 4200.

Jan van Driel

[scalbers]

Animation between Sols 4193 and 4197 using official mosaics. Charting a new course to the west and moving further into the channel.



Longer steppable drive animation with a combination of official and Jan's mosaics from Sol 3435 up to the present.

[fredk]

The accuracy of the colours here isn't very certain, but I don't recall anything showing up this vivid a shade of green with these cameras before:

[Floyd]

Did a wheel go over this? Any one want to speculate as to what mineral would have this appearance?

[fredk]

Yes, they drove over it and cracked it open. You can see it on the left of this navcam frame:
https://mars.nasa.gov/msl-raw-images/proj/m...NCAM00200M_.JPG

[neo56]

The accuracy of the colours here isn't very certain, but I don't recall anything showing up this vivid a shade of green with these cameras before:

I think part of the green tint is due to debayering artifacts. Here is what I obtain after removing them.



Here is also the MC100 context mosaic of the workspace.

[fredk]

I think part of the green tint is due to debayering artifacts. Here is what I obtain after removing them.

It's common to see green blocks when doing a simple deBayering. But they appear on smooth areas, where the jpeg compression algorithm is tricked into compressing too much. I don't recall green artifacts on irregular areas before. Also the green deBayering artifacts follow the 8x8 pixel DCT jpeg block pattern, which these greenish areas don't.

How do you remove the green blocks - is it possible you've removed too much green?

The overall colour balance may be off somewhat in my image. And there certainly might be something about the rock here that causes some new type of artifact - we may have to wait to see colour images from the original data to know for sure.

[serpens]

Being Mars, the automatic reaction is that it's calcium sulphate.

[neville thompson]

Gigapan - MSL 4202 MR
© NASA/JPL-Caltech/MSSS/NeV-T
--------------------------------------
GigaMacro - MSL 4202 MR
© NASA/JPL-Caltech/MSSS/ASU/NeV-T

[Floyd]

re fredk and serpens
Gypsum can be green if it contains a little copper.

[neville thompson]

It's common to see green blocks when doing a simple deBayering. But they appear on smooth areas, where the jpeg compression algorithm is tricked into compressing too much. I don't recall green artifacts on irregular areas before. Also the green deBayering artifacts follow the 8x8 pixel DCT jpeg block pattern, which these greenish areas don't.

How do you remove the green blocks - is it possible you've removed too much green?

The overall colour balance may be off somewhat in my image. And there certainly might be something about the rock here that causes some new type of artifact - we may have to wait to see colour images from the original data to know for sure.

Try using HLVG been using it for years - pretty good

https://www.deepskycolors.com/tools-tutorials/hlvg/

[neo56]

It's common to see green blocks when doing a simple deBayering. But they appear on smooth areas, where the jpeg compression algorithm is tricked into compressing too much. I don't recall green artifacts on irregular areas before. Also the green deBayering artifacts follow the 8x8 pixel DCT jpeg block pattern, which these greenish areas don't.

How do you remove the green blocks - is it possible you've removed too much green?

The overall colour balance may be off somewhat in my image. And there certainly might be something about the rock here that causes some new type of artifact - we may have to wait to see colour images from the original data to know for sure.

I use "Replace color" tool on Photoshop, selecting greenish areas and decreasing their hue of -10.
It's possible I've removed too much green. But I suspect the color of these bright crystals being affected by some sort of green artifacts.
The correction made by Neville is really good at removing greenish blocks.
Let's see what the color calibrated data look like when it'll be released.