Phoenix PDS Releases Options

[djellison]

" The actual delivery schedule will exceed these requirements: the Phoenix Project will make at least two deliveries to the PDS, the first one no later than six months after Sol 30 data are received on Earth, and the second one no later than six months after Sol 90 data are received on Earth"
http://pds-geosciences.wustl.edu/missions/phoenix/index.htm

Due, I would have thought, in the not too distant future. We are 161 days after Sol 30, which was June 25th. So according to the PHX archive plan - it'll be Chrismas day

Doug

[Bjorn Jonsson]

Then I know what I'll be doing on Christmas day if the archive plan is followed: Testing to see if IMG2PNG handles the Phoenix data properly or if some modifications are required .

[slinted]

Looks like they will be creating an Analyst's Notebook interface for Phoenix like they did for MER. According to the placeholder, Dec 24th will be the release date. Santa's coming early!

[elakdawalla]

Peter Smith mentioned the Dec 24 release date of the first 30 sols of data at AGU yesterday, so they're serious about that. He mentioned a date for the rest of the data but unfortunately I missed it.

--E

[OKB001]

Oh that is good news! Thanks for the head-ups djellison!

[peter59]

Christmas surprise !

http://pds-imaging.jpl.nasa.gov/data/phoenix/
.. and Merry Christmas to all.

[tedstryk]

Christmas surprise !

http://pds-imaging.jpl.nasa.gov/data/phoenix/
.. and Merry Christmas to all.

And to all a good night...Sweet!

[peter59]

I like to view the horizon. The difference in quality between the images in JPG format and radiometrically corrected images in IMG format is awesome. I am delighted that Phoenix landed in the vicinity of such high hills.
compressed JPG:
http://www.met.tamu.edu/mars/i/SS018EFF897...5_11E1EL1M1.jpg
http://www.met.tamu.edu/mars/i/SS018EFF897...3_11E1EL1M1.jpg
http://www.met.tamu.edu/mars/i/SS018EFF897...8_11E1EL1M1.jpg
radiometrically corrected IMG:
http://pds-imaging.jpl.nasa.gov/data/phoen...5_11e1el1t1.img
http://pds-imaging.jpl.nasa.gov/data/phoen...3_11e1el1t1.img
http://pds-imaging.jpl.nasa.gov/data/phoen...8_11e1el1t1.img

[elakdawalla]

Thanks for that handy illustration, Peter!

--Emily

[djellison]

Using RAD's from http://pds-imaging.jpl.nasa.gov/data/phoen.../browse/sol018/ (the phxssi_0xxx isn't, as far as I know, radiometrically calibrated)

Then using IMG2PNG (yes, it works !! ) with img2png *.img -r -s6

Then doing the downsampled R A,B,C - as a colour layer over the full res R1.

Result.

Attached thumbnail(s)

 

[TheChemist]

What a wonderful and eerie view !!!

[tedstryk]

Wow, impressive! I have wanted to have a super-res go at those hills since I first saw them. I hope there is a proper sequence. If not, I hope to cobble one together. Now if the darn data would download (this is going painfully slow - I started at 9:00, and I am not half way).

[tedstryk]

What is the difference between SSI-1 and SSI-0 (SSI-0 hasn't started coming down yet or I would look myself)?

[djellison]

When I used IMG2PNG the '1' stuff mosaics perfectly - it seems to be a true equivalent of the MER RAD images (as it should - I'm guessing the processing pipeline is essentially the same). The '0' stuff doesn't. Some frames match, some frames don't. I think the numbers reflect the level of processing.

The Level 1 folders also include a lot of extra processed image types - but I like to stick to the orig. RAD frames.

Doug

[tedstryk]

Thanks. I was trying to figure out just what was in the zero folders. I don't know what the issue is. I have also downloaded a bunch of MER and Cassini stuff today, and it zipped right down. The Phoenix stuff is taking eons.

[fredk]

I like to view the horizon.

Me too. It sure is good to see cleaner versions of these images. Here's one of my favourite views:

And in this view I thought there was detail visible on the distant hills from the jpegs. Now there's no doubt that those details are real:

[djellison]

Sol 2 mosaic - Just R A,B,C - very red. I think I prefer R ABC x R1.

Attached thumbnail(s)

 

[ugordan]

Wow, that is red. I think it has to do with the fact the data is linear, not gamma-corrected for computer screens. It looks less garish with a correction, but could still probably be tweaked to fudge filter bandpasses better into sRGB colorspace:



At least the color chips look OK!
The RB filter is furthest from the sRGB colorspace so mixing a bit of red into it gives this (80% green, 20% red):



Looks a bit better, doesn't it?

[brellis]

What a treat! My jaw drops all over again. Thanks to all for sharing your work and comments on the images. It really helps me get a feel for what it would look like for a person standing there on Mars.

[Stu]

Hmmm, I kind of like the really red one better...

[djellison]

I love data. Sols 15 and 16 chunks of the Peter Pan

Attached thumbnail(s)

 

[OKB001]

I love that you love data Please keep the love flowing

[djellison]

And the Sol 30 Solstice image - not best pleased with colour on this one. I'll leave processing the full peter-pan to James

Attached thumbnail(s)

 

[Deimos]

This is fun.

To answer an earlier question, the phxssi_0xxx should be raw DN, uncorrected for anything other than decommutation (not packets, not compressed anymore). The phxssi_1xxx is processed using JPL's version of the pre-landing calibration. The phxsci_0xxx should be the best for all cameras. The difference for SSI should be relatively small (better bias correction, better flat field, different smear removal for certain binned images). Ironically, half my software choked on the data for this release, but is mostly back up. So if there are cases where phxssi_1xxx is better than phxsci_0xxx, I'd appreciate examples.

Personally, I like the ones where the filter color-space is mapped to sRGB, rather than straight substitution. YMMV.

Cheers.

[djellison]

I like the ones where the filter color-space is mapped to sRGB,

That's where Dan Crotty comes in - he's good at that

Dan - get to it !

Doug

[slinted]

Working on it I should have something to share by next week.

[peter59]

These distant hills fascinated me. I tried to identify them on Mars Reconnaissance Orbiter images. I'm not sure if my identification is correct. What do you think?

Map

MRO - PSP_007642_2480, PSP_006996_2480, PSP_007497_2480


left hill - PSP_006996_2480, right hill - PSP_006996_2480

[Doc]

I thought those hills correspond to the Heimdell crater passed by Phoenix during EDL phase. Now I'm not so sure!

BTW all of Phx's data is now available. Now its time to play around with the OM pics

[peter59]

Now I have no doubt as to the identification of the image PSP_006996_2480.


Other photo (PSP_007642_2480) taken in early spring or late autumn presents the hill with frost in the bottom of the slope.

I regret that Phoenix could not be active at a time when frost appears.

[fredk]

Peter, your identification agrees with that of Phil Stooke, and that's good enough for me! (Those two hills are labelled A and B in Phil's map.)

We can see a bit of the outskirts of Heimdall crater rising up towards the NE.

[djellison]

Nice to see the tell-tale data out

A few graphs plotting direction, speed, local time etc etc. What's obvious is Northern and Southern winds being the strongest, with little from the East and West.

Attached thumbnail(s)

 

[Doc]

Cool graphs Doug.
I went through the OM data (analyst helps a lot; you can't get lost in the data) and got to make this nice one form sol 19



I think thats a sample from Mama bear if not mistaken; those must be magnetite rich grains collected at the magnets.

[slinted]

I'm slowly starting to get a handle on the calibrated color for Phoenix from the PDS. The science RDR versions of the RAD images look fantastic!

I've got a gallery up with my progress so far: Phoenix Calibrated Color

It currently has the SSI right-eye color images for the following filter combos (in order of preference): 1ABC2, 1ABC, ABC, 1BC, 1B2 and 8B2. The end of the filenames list the filters used. The three- and four-filter sets use color formulas fit to the expected output color based on a five-filter training set. Each frame is coregistered to correct for the shift that sometimes occurred between filters. There is not any scene-based chromatic adaptation being used, so the hues are "as they are". The brightness is held constant throughout all the images.

Now for the caveats! I haven't yet put in pan sharpening to make good use of the mixed high-res low-res images, so the bulk of the panorama frames are a bit fuzzy at full zoom. Also, a few of the images have some visible artifacts from saturation (sol 16, and parts of the deck on sol 21 and 30).

As an example image, here's a look at Dodo-Goldilocks on sol 18:

[ugordan]

I've got a gallery up with my progress so far: Phoenix Calibrated Color

Looking great, slinted! Very consistent colors.

[4th rock from th...]

Excellent work Slinted.

Very consistent and it highlights some interesting sol by sol variations of contrast.
The images also remind me of Mars Pathfinder (the filter set is close i think).
Perhaps any chance of having a calibrated MPF gallery on your site ?

[peter59]

Sol 31 - 90 released !

http://pds-imaging.jpl.nasa.gov/data/phoenix/

[Deimos]

In anticipation of the next PDS release, I'll share some images that are available but have been unadvertised. They are calibrated SSI color images, processed to reflectance (ie, corrected to the illumination present) and then to approximate true color. The processing was done for any image set that supported it, using semi-intelligent interpolation for data with few filters. So, left eye color is available to go with really good right eye color for Happily Ever After images (id: 165C) for cross-eye stereo types (although the directories were not organized for such types). As a bonus, wherever compression levels allowed, deconvolution was performed on the input images to the color processing. And, again when conditions allowed, missing packets were interpolated over (but frequently they still show up as an off-color band).

The images are at : http://www.met.tamu.edu/mars/color.html

Most or all of these images will also be available through analyst's notebook when the release is out.

Mark

[Stu]

Wow... thanks Mark, those "conoisseur" images are absolutely gorgeous...!

I miss Phoenix

[peter59]

Sol 91 - 151 released !

http://pds-imaging.jpl.nasa.gov/data/phoenix/

[Guest_Bobby_*]

I miss Phoenix also Stu.

Even know it wasn't a rover. It was an amazing lander to us Mars Fans

A Question. When did Mars winter for Phoenix end and I know the odds of Phoenix
coming back to life is unlikely. If it did come back to life. When would it have and has
JPL made any attempt to communicate with Phoenix???

I think a part of all of us would love to see this happen but deep down we know it won't.

At least my name is on The DVD on Phoenix so I know a part of me even in Binary Code
will be part of The Rover.

Bobby (The Quiet Mars Fan)

[helvick]

It's just approaching the middle of winter for Phoenix right now. If the worst case projections were correct she's sitting encased in CO2 at the moment which is quite a surreal thought. At a guess _if_ she survived then it would be worth starting to listen for signs of resurrection towards the end of this [earth] year\ early 2010.

[marsophile]

I haven't been able to find the Atomic Force Microscope (AFM) images in this release. Can someone provide a pointer? Are the AFM images not going to be put into the PDS at all, or are they just not out yet?

[djellison]

They're in there - use the analyst's notebook to find them. They're not in an easily consumable format, however.

[marsophile]

The AFM archive in the PDS is not yet complete. The EDRs are there, but the RDRs have not been deposited yet. The last announcement said the RDRs would be delivered later, at approximately June 1, but it is now June 8, and still no sign of them. Why the holdup for the RDRs?

[djellison]

Why the holdup for the RDRs?

Why not email the appropriate PI as listed in the PDS documentation.