I'm sure I asked this somewhere else at some point..I've searched but no luck - anyway

I've got all the Pancam data up to 270, and am DVDiffying the lot.

HOWEVER

Has anyone found an FTP server that has the Navcam and Hazcam RAD's? I have the MI rad's coming down at the moment - but cant find Nav/Hazcam anywhere. I dont think I could handle the pain of downloading them individually ;P

Doug

I have now uploaded a new version of my IMG2PNG utility mentioned earlier in this thread (on January 16, 2005 to be exact). This new version correctly handles 16 bit PDS Cassini image files. They were incorrectly handled in some cases by the earlier version, including the case where a detached label file is present in the same directory as the IMG file.

It can be downloaded at http://www.mmedia.is/bjj/misc/img2png.zip

This new version still does not perform any Cassini-related calibrations since the Cassini calibration DVDs still haven't appeared at the PDS.

A new version of my IMG2PNG utility:

http://www.mmedia.is/bjj/utils/img2png/

The main change is the ability to calibrate Cassini images when converting them to PNGs. I also fixed a crash involving Stardust images.

Quick thought - you mentioned some time ago that the option to pull the headers out as a text file - did you ever do that? Not a biggie, but it would be quite usefull now I'm trying to orient terrain wedges

Doug

I had forgotten about this but it is trivial. Actually I have a new version ready with several bugfixes that I was about to release. I'll add this capability before releasing the new version, possibly this weekend.

Speaking of IMG2PNG I have for some time wanted to add the conversion of FITS to PNG but haven't had the time to do so yet. I will probably rewrite most of my PDS file reading code from scratch first. It was originally written back in 1994 (!) to read Voyager images, was one of my very first programs written in C++ and is therefore, frankly speaking, badly written and very difficult to maintain.

And speaking of new features, I recently added calibration of Galileo images to IMG2PNG. However, this required me to process the Galileo calibration files into a special format and they are too big (55 MB zipped) to release together with IMG2PNG .

Wow - that was quick - top stuff - the Fits addition is going to be very cool w.r.t. NEAR stuff

Doug

The headers are in text format, so you can actually open .IMG file in WordPad and read the header.

I have now uploaded a new version of IMG2PNG.

This new version has the ability to dump some of the image labels to text files and it is possible to tell it not to output any PNG files (can be useful to check for bad input files or to create thumbnails only before converting).

It can be downloaded at http://www.mmedia.is/bjj/utils/img2png/

Fire up those FTP clients, and dust off your wgets, MER PDS 8 (sols 540-630) was released today!

Excellent! Let us know when the latest callibrated color images are online.

I've posted up color images from PDS release 8 at http://www.lyle.org/~markoff/ or in Midnight Mars Browser under Update:Advanced update images:update images from: lyle.org, sols 541-630. I still haven't finished my work on dynamic whitepoint, so this release has been processed with the same fixed whitepoints as my last release, which are more appropriate for the dusty conditions seen on these sols.

Top stuff Dan - much appreciated.

Doug

They look terrific. Thanks

Nico

Looks great. I will try to update the MMB panorama metadata later today to include the new stuff.

The MER EDLT document has been moved to:

http://library-dspace.larc.nasa.gov/dspace...a-2004-5092.pdf


No, the implication is that the reconstruction is difficult, and incomplete and may remain so because there was not enough instrumentation to constrain the descent parameters.

Yes, the RAD firing reconstruction columns are filled out quite well, as are many of the Parachute deployment parameters, but such critical parameters as the altitude at which the lander initiation sequence began, and the exact time of the landing are not in the tables. Bruce has indicated that at least for Spirit, the lander initiation was perilously close to the ground – or something – I can’t see anything in the reconstructions that would indicate the MER-1 descent was significantly different from that of MER-2.

The important thing is that the need for more complete instrumentation (pressure, density & temperature) during the descent & landing phase has been recognized and is being engineered into future landers.

MMB panorama metadata has been updated for the latest callibrated color images from PDS 8.

Latest update, up to sol 720, is available at MER Analyst's Notebook

For Spirit, that includes the descent down Husband Hill, "Seminole" Pan, "Comanche", and "El Dorado".

For Opportunity, that's mainly Erebus (the extended stay at Olympia).

I hope slinted is in town!

The ftp server at pds-geosciences.wustl.edu apparantly had a bit of a hiccup with this release. Opportunity was available last Friday, but the Spirit RADs were made available today. I'll be getting my release together in the next couple days. Also in this release is the Erebus pan and a couple nice looking sunsets from Opportunity.

Late is better than never right?

The color images from PDS 9 are now available at http://www.lyle.org/~markoff/ or in Midnight Mars Browser under Update:Advanced update images:update images from: lyle.org, sols 631-720.

For this release, I lowered the max brightness of Spirit's image by 25% (to 0.09) to get a little more detail out of the images. Opportunity is unchanged.

Some highlights include:


Phobos/Deimos nighttime color image (I'm still trying to make a real color image out of this one, but the smear from the longer exposed frames is problematic).


Comanche


El Dorado

Just two quick stitches because I'm too lazy to do better 5 minutes before I go to bed:

http://paranoid.dechengst.nl/mars/Comanche.jpg (535KB)
http://paranoid.dechengst.nl/mars/Somehill.jpg (670KB)

Great work - some of that Commanche imagery is just sublime.

Doug

Some? Earlier this evening I spent the best part of an hour drooling over the new images. Just beautiful, beautiful visions of Mars.

Said it before, but worth repeating: your work doesn't get nearly enough exposure or credit. Looking at some of those pictures I feel like I could reach out and touch those rocks...

Indeed. I'm downloading them through MMB right now, and although I planned to do some stuff around the house, it's just too mesmerizing watching them appear one after the other. These are gorgeous.

Those are indeed some very nice pics; they give a real "your are there" feeling.

About the radiometric calibration ... and this may well have been asked and answered before ... I have been watching the color calibration targets becoming increasingly dusty over time, and wondering how the color calibration is managed with a dust covered target? Is there some trick of matching exposures to those from early in the mission when they were clean, and then somehow subtracting off the difference? Always curious to learn a nuance of image processing

IIRC, as far as the Pancam team goes this was talked about extensively in the most recent "Jim & Doug Show". How slinted adjusts his technique for changing tau is no doubt an equally fascinating story (which I probably won't understand either).

Still watching the images download. Still being blown away.

I've updated the MMB metadata with Daniel's new images, I think. Here are a couple preview views.

Here is a large combined MMB/Autostitch view of El Dorado:

I posted a QTVR "El Dorado" panorama with Daniel's color images here. I think you really have to see it in perspective projection to fully appreciate the view. Unfortunately this is less than half full Pancam resolution, due to technical limitations on my side at the moment.

Great work from both of you - it's a 4-times-a-year-highlight flicking through Dan's new release in MMB

Doug

I gave the "Jim & Doug Show" a second listen - I'd previously been distracted during that section. The technique described is about what I had expected. Good to know

Awsome work, Micheal, Daniel ... I love it !

Yesssss!! Very good! I like the color and the light ambiancy, particulary on the Oppy pic.
Thanks!

Here you go: "Erebus Pan" using Daniel's callibrated color images.



10000x5000 pixel version

I have to clear off some hard drive space before I can generate the full-size version

very beautiful ,

but why this color difference with nasa version ... ?

http://marsrovers.jpl.nasa.gov/gallery/pre...-B652R1_br2.jpg

Maybe Daniel can comment on that if he stops by. I'm using his callibrated color images, not JPL's; they obviously use different techniques. I like Daniel's but of course, JPL's are nice, too. Very hard for me to speculate which is more 'accurate', if there even is such a standard.

Edit: I think that may be an older version of the pan, with fewer filters, for one thing. I don't see the RAT hole in that one.

Seminole pan:



10000x5000 pixel version

Sorry for being a month and a half late on this reply, but I did want to (eventually) give a reasonable response.

First off, I think the better starting question is this: why does their Erebus pan look different than the other Opportunity panoramas? Scroll down through all the panoramas on this page and it becomes apparent that something is fundamentally different about the Erebus panorama. We're looking at the same blueberries, the same dust, and the same color outcrops as all the other Opportunity panoramas, but the Erebus pan gives a different color impression of all 3. The time of year was different, as was the dust load in the atmosphere, but I'm not sure this completely describes the difference we are seeing. I don't know this for sure, but I think the issues lies in the filter combination. L2,L5,L6 was the standard filter combo for Lion King, Endurance South, and Rub al Khali. Since there are many L256 panoramas and mosaics taken by Opportunity, I'm sure they have spent a good bit of time fitting the data from these 3 filters to the more accurate 6 filter images. The Erebus pan used L2, L5, L7. A better match between the all-6 filter images and that particular combo could have been made, but I think we can all forgive them for getting close but not quite on the mark with this particular panorama. For more information about the Cornell teams methods for true color generation and fitting 3 filter to 6 filter, see Bell III, J.F., J.R. Joseph, J. Sohl-Dickstein, H. Arneson, M. Johnson, M. Lemmon, and D. Savransky, In-Flight Calibration of the Mars Exploration Rover Panoramic Camera Instrument.

While I do think the filter combo of the Erebus pan is the biggest reason for the deviation, there is another difference between my processing pipeline and the one used by the Cornell team which differentiates my images and theirs. The input data is the same (I'm using their calibrated files) as are most all the steps in processing. We do differ in our processing mostly in the final steps, or more specifically in how far we process the images. The Cornell team finds the color of each pixel based on the spectrum, then display those colors. I find the color at each pixel, then adapt the entire scene based on the content, in an attempt to reproduce the effect that standing in that scene would have on the human eye (chromatic adaptation). In that one extra step, my processing deviates significantly from the Cornell teams' work. While my methods for accomplishing this goal may be a bit un-orthodox, taking chromatic adaptation into account is standard in most terrestrial photography. In fact, most digital cameras do it automatically. Although I'm not sure why the Cornell team chooses not to include this step, I think the omission is understandable from their perspective. They are geologists, not neurobiologists. Their primary concern is the spectrum of the scene, not the (imperfect) way humans would perceive it. Of course, I would be happy to see them address this issue in the future, but I can understand their lack of concern for it in an immediate sense.

yes is the good question

I was astonished initially ,the image is rather "green" and sky too

thank you for your answer

Xav.

Latest update, up to sol 810 is at MER analyst's notebook. Homeplate images

Ahh - that's better - a not-quite-so-warped Gibson pan.

Doug

i dont know is the good post to make a gallery ... but i follow ...

Spirit Site AODQ

Click to view attachment

and vertical projection at the same site

0.5cm/pixel

Click to view attachment

SPIRIT Site APA0

Click to view attachment

Very nice, Indian. (Heavily compressed, obviously.)

The color images from PDS 10 are now available at http://www.lyle.org/~markoff/ or in Midnight Mars Browser under Update:Advanced update images:update images from: lyle.org, sols 721-810.

Even though there are fewer color images in this release than any previous one, there were some gems in this group, mostly from Spirit.

Arad:




Tyrone: (navcam + pancam, with enhanced contrast)

I've update the MMB metadata.



large version

The latest update, to sol 900, has been added to MER Analyst's Notebook.

It'll be a couple days yet before I have the color from this release finished. Back around sol 450 I adjusted the chromatic adaptation to compensate for the dustier conditions during southern spring/summer. That adjustment has now become inappropriate since the images in this PDS release were taken during the southern winter and from the looks of the skies, it was a relatively dust free winter.
So I'll be shifting the white point back towards more neutral lighting. I'll probably reprocess the last release as well (the blue rocks were a pretty good indication to me that something was off in the last release, but I didn't have the time to make the adjustment then.)

I've updated my site to include the color images from PDS 11 (sol 811-900) using a new form of processing. PDS 10 (sols 721-810) has also been reprocessed using the new technique. Both are now available at http://www.lyle.org/~markoff/ or in Midnight Mars Browser under Update:Advanced update images:update images from: lyle.org, sols 721-900.

The problem I have been trying to address is the brightness scaling. Previously, I was producing color images on a fixed and absolute scale of brightness (a 0 value in the color images represented 0 radiance). This severely limited the available range for the actual content of the scene, since brightness of shadows and the sunlight sections of images are not that far removed on Mars (indirect light plays a significant role in lighting all scenes). Basically, the absolute scale caused the images to be very bland and without contrast.

To fix this, without distorting the hue, I've settled on a custom luminance compression based on the L channel of the L a*b* color space (the same used by PS for 'contrast' stretches). This allows the shadowed areas to be positioned in a darker part of the image range, leaving the rest of the range available for the sunlit portions of the scene. It is a somewhat arbitrary decision, but in doing so I am trying to match the natural adjustment human eyes+brain would make when looking at these scenes. In terms of 'true color' or not, I think this change is a matter of choosing what which parts of perception to get right when having to compress the real scene into a monitor display. I could get the absolute scale "correct" but in doing so sacrifice the natural contrast of the scene, or as I am now doing, preserve the contrast that would be perceived while giving up the absolute scale. The chromatic adaptation issue is much the same, do I try to get the absolute color (like the Pancam true color) correct or do I try to convey the way these colors would be perceived once your eyes adjusted to the scene in person? I have chosen to try achieve correct perception as the priority.

Since this is a compromise between true color and true perception, there are some problems at the edges of the range. Some shadows do get clipped to black, and some parts (particularly high reflectance bits of the rover itself) are being clipped to white. This is downside of giving contrast-potential to the bulk of the scene.

I hope to eventually apply this to all the rover images going all the way back to the beginning of the mission, but I need to settle on how to dynamically adjust the chromatic adaptation based on changing dust levels first, so it may be a ways off.