stim3on

stim3on, 8 days ago

@65dBnoise is that one from Sol 1155 or did we get new rover deck images?

stim3on, 9 days ago

@65dBnoise Great job processing these SuperCam images by the way! Your flatfielding is top notch and so is the contrast processing!
I'm really struggling to produce good colors with my process and I don't know what I'm doing wrong, they all come out with a strange green cast that doesn't go away with white-balancing 😬

I would really appreciate if you can share your workflow so I can compare it with mine. :)

stim3on, 8 days ago

@65dBnoise Thanks a lot, I will take a look at how the results from your process compare to mine.
I made my own flat from possibly the same images and do the processing in python.
It's very possible that I made a mistake in my math, possibly the linearization of the brightness values that is causing my problems.

Am I correct to assume you are using the pre-debayered EBY images and not the ECMs?

stim3on, 8 days ago

@65dBnoise I'm actually using the same sol 868 images as flats. These are called skyflats (self explanatory) and are regularly taken by the science cameras for in-flight calibration.
They are usually taken in pairs of two with reverse orientation so local sky gradients cancel out after stacking (If I remember correctly at least)

Here is the histogram of my flat - it's in linear brightness values, so no gamma applied like in yours.

stim3on, 8 days ago

@65dBnoise This is what my stacked flat looks like. I also subtracted the mean value of the shadowed areas to remove the sensor bias offset - however this is under the likely wrong assumption that no stray light enters this part of the sensor. This is one of the possibilities where I might have made a mistake.

The flat was computed from the PDS EDR data (linarized, not debayered)

stim3on, 8 days ago

@65dBnoise By the way, I usually try to avoid using the EBY images since they are still affected by a small stretch on the high and low end of the brightness range. This can affect color accuracy, so I always use the ECMs and debayer myself when possible.

video/mp4

stim3on, 8 days ago

@65dBnoise Ah, no I didn't actually desaturate the image, this is still the raw bayer pattern you're seeing. I just rescaled the three different pixel types to approximately be white. Debayering happens much later in my processing and is not something I do in my python processing. Rather, I offload it to Lightroom which has better debayering algorithms which also allow for some other fancy processing.

stim3on, 8 days ago

@65dBnoise I need to give it a try first. Though after some more experimentation it seems like my pipelines results aren't actually that different from yours.
I just had some suboptimal settings in Lightroom that made the colors look strange.

I think the image is actually quite satisfactory now.

I'm still seeing a visible haze on the edge of the FOV, but I know that the SuperCam team is already investigating this as stray light effects.
https://www.hou.usra.edu/meetings/lpsc2024/pdf/1318.pdf

stim3on, 8 days ago

@65dBnoise If you would like to try my flat as well, you can download it here: https://www.pastefile.com/vcehqs

remember that it is still in it's monochrome bayered state, but debayering might not actually do much different since it's visually quite uniform.

It's also in a linear brightness range, so if you want to apply it to the ECM/EBY images directly you need to apply a gamma of 2.2 first (e.g. via levels in Gimp)

stim3on, 8 days ago

@65dBnoise Haha Cheers! Good luck with it ;)

After some more manual adjustments to compensate the halos I was able to seamlessly merge the SuperCam images into this Mastcam-Z 110mm image:
https://flic.kr/p/2pUujxN

stim3on, 10 days ago

@65dBnoise A 360 Navcam panorama might soon be coming to a mastodon instance near you :)

stim3on, 11 days ago

This post by @65dBnoise has a map of the current location of the rover next to this hill.
The rover was actually on track towards the west, so apparently something about this hill peaked the interest of the science team.

https://mastodon.social/@65dBnoise/112517447001873649

stim3on, 11 days ago

@tsturm Here is a tighter crop on this part of the image. I'm not sure if the dark spots are pores or darker grains in the rock. I'm thinking the latter, similar to how granite or gneiss can look on Earth, but the resolution isn't good enough to say for sure.

stim3on, 11 days ago

@TMEubanks @tsturm I don't see any recent impact crater in the MRO HiRISE imagery.
And looking at the other dark rocks they also seem to have a very clean side (facing the rover) while the other sides are much more covered with dust.
I could imagine some local wind effects keeping these sides much cleaner than usual.

stim3on, 11 days ago

@TMEubanks @tsturm You can see this dust accumulation difference in the wider Navcam image
https://mastodon.social/@65dBnoise/112517675126609898

stim3on, 11 days ago

Here is an alternative processing with exaggerated color saturation to highlight the differences of rock colors.

Credit: NASA/JPL-Caltech/ASU/Simeon Schmauß

stim3on, 11 days ago

This is a wider view captured by the right Navcam of the rover, processed to a natural color apperance.

Note the orientation dependent amount of dust accumulation on these boulders. The side facing the rover is almost free of dust while the top isn't. Could this be due to local wind effects in Neretva Vallis?

Follow the link for a 170MPx version of this Navcam image fused with the higher resolution Mastcam-Z images.
https://www.gigapan.com/gigapans/234902

Credit: NASA/JPL-Caltech/Simeon Schmauß

stim3on, 11 days ago

@TMEubanks @tsturm
Well, then take a look at the dark rocks in the foreground where we can also see the top and backside. They are much more dusty than the rover facing side.
If dust accumulation was equal on all sides, these rocks wouldn't look like that and all the other dark rocks would look much dustier too.

stim3on, 11 days ago

@65dBnoise that white one already caught my eye in the Navcam images, I'm curious what it's made of

stim3on, 11 days ago

@sharponlooker @65dBnoise
It's in the works, and I definitely agree it's one of the more interesting panoramas captured by the rover thus far.
I'm very glad they decided to capture some additional images of the rover for this one, which usually doesn't happen with Mastcam-Z mosaics.
It will still be a ton of work to process it to the level of quality I strive for, and I'm also pretty busy at the moment but it's slowly coming together :)

stim3on, 26 days ago

@65dBnoise Yeah, good news I guess 👍

stim3on, 26 days ago

@65dBnoise @AkaSci Hmm loking at it in motion in @Landru79 's gif, it sure looks to me like physical arm motion with all the jittering.
https://x.com/landru79/status/1790059963923996821

stim3on, 21 days ago

@65dBnoise I bet there is an interesting story to be told about a group of engineers in the Marsyard working long nights to develop a plan how to regain focusing capability with Sherloc by incrementally moving the robotic arm :)