Printable Version of Topic

Posted by: tedstryk Jan 20 2012, 08:45 PM

It appears Rosetta stuff, including the Mars flyby Osiris data, has hit the PDS. I don't have time to look at it now, but I'm downloading it. http://pdssbn.astro.umd.edu/holdings/

Posted by: ugordan Jan 21 2012, 12:15 AM

Finally...

A couple of RGB products:
NAC view of Moon from 2nd Earth flyby, magnified 2x:
http://www.flickr.com/photos/ugordan/6733334533/sizes/o/in/photostream/

Full frame NAC view of Mars with Phobos and its shadow:
http://www.flickr.com/photos/ugordan/6733326721/sizes/o/in/photostream/

Posted by: stevesliva Jan 21 2012, 12:23 AM

^ I can't believe they just sat on that Mars shot for nearly five years.

Posted by: elakdawalla Jan 21 2012, 01:23 AM

Is this a crescent view of Deimos???? That huge south polar indentation identifies it, I think. Dang, should have checked the index file, it's Phobos. (Thanks, Gordan.)

Nearly every shot of Mars includes Phobos.

And to think these weren't shared with the world because they were "overexposed" and therefore not of any interest.......grumblegrumblegrumble......

Posted by: elakdawalla Jan 21 2012, 01:27 AM

OMG....is this what I think it is??? Can it be one of Mars' moons, mostly Marslit, setting behind Mars' nightside, Mars' atmosphere forward-scattering sunlight to us...? There is a TON of these, a whole animation of crescent Phobos? setting. Not sure what the source of the noise is. If my interpretation is correct, WOW.

Posted by: ugordan Jan 21 2012, 01:29 AM

The index file labels that as "Post CA Phobos Spectrophotometric". Here it is in RGB, magnified 2x:



Note that all images are mirrored so they need to be flipped. I have no idea which way north is supposed to be.

Posted by: ugordan Jan 21 2012, 02:13 AM

Global inbound shot revisited:

http://www.flickr.com/photos/ugordan/6733745309/sizes/o/in/photostream/

Posted by: SFJCody Jan 21 2012, 04:12 AM

Lovely. I'd like to see this or something very much like this used as the 'default' Mars image rather than that Viking photomosaic centered on Valles Marineris.

Posted by: tedstryk Jan 21 2012, 04:14 AM

Gordan, lovely treat. Just got home from the symphony. I will have to have a crack at these in the coming days.

Posted by: stevesliva Jan 21 2012, 04:22 AM

Love the way it gets whitish at either edge. Just beautiful.

Posted by: Phil Stooke Jan 21 2012, 05:01 AM

Fantastic! Thanks everyone.

The thing that looks like an earwig's posterior in the Mars shots is Gale crater.

Phil

Posted by: ugordan Jan 21 2012, 11:13 AM

Thanks for pointing that out Phil, that's one more note for the images.

Posted by: Stu Jan 21 2012, 11:47 AM

At the time of the encounter it was made known that some images of a "crescent Mars" had been taken, but hadn't been released because they were "over-exposed". I was - some may recall! - rather upset about that, and emailed the ROSETTA team more than once asking for the images to be released, even if they were over-exposed, because they were still of interest. Nope.

Found them.

Fair enough.

Posted by: Paolo Jan 21 2012, 01:52 PM

I had just a quick look at the data, but I couldn't find what I was looking for: the raw Philae image of the orbiter with the background of Mars (to me, the best UMSF image ever!)
has anybody seen it?

Posted by: machi Jan 21 2012, 02:31 PM

I think, that it's similar situation as in case of the Stardust' Tempel 1 images. Mars and Earth data from Rosetta are not officially released and same situation in case of the Philae lander.

Nice findings!

Ugordan:

Beautiful images!

Posted by: ugordan Jan 21 2012, 02:42 PM

Thanks, machi. Here's another one with Phobos:

http://www.flickr.com/photos/ugordan/6736094773/sizes/o/in/photostream/

I like the quality of the NAC images. Too bad most of the stuff from Mars and Earth flyby 2 is overexposed and unusable.

Posted by: machi Jan 21 2012, 02:50 PM

I was wrong, evidently Mars and Earth (two flybys) data are oficially released - http://pdssbn.astro.umd.edu/data_sb/missions/rosetta/index.shtml
But still we must wait for data from the lander.

Posted by: john_s Jan 21 2012, 05:24 PM

I love the details in the clouds along the northern terminator in that one...

John

Posted by: Phil Stooke Jan 21 2012, 06:01 PM

Yes, the clouds are exquisite! Good work everybody.

Phil

Posted by: tedstryk Jan 21 2012, 09:58 PM

Here is a pretty one.

Also, the Phobos images are conjuring memories of 1969.

Posted by: Phil Stooke Jan 21 2012, 11:30 PM

Very nice, Ted, as usual... but flipped E-W (as you probably knew)

Phil

Posted by: Stu Jan 22 2012, 12:02 AM

Bit late to the party...

Posted by: tedstryk Jan 22 2012, 01:00 AM

I realized it when I saw it next to the Mariner-7 Phobos over Mars image. Oops.

Posted by: djellison Jan 22 2012, 03:48 AM

Are we (apart from UG) using img2png? I'm getting a combo of very very dark moon images from the Earth flybys, but full pixels in some of the Mars images.


D

Posted by: elakdawalla Jan 23 2012, 11:48 PM

I'm using IMG2PNG, and getting nice results. What do you mean by full pixels?

Here's my take on the inbound global shot.

Posted by: djellison Jan 23 2012, 11:51 PM

Saturated white pixels - meaning I can't make a shot like that one, as the polar cap is bleached out.

Did you just do img2png.exe *.img - or did you use any tags on the end?

D

Posted by: elakdawalla Jan 23 2012, 11:56 PM

Nope, just vanilla IMG2PNG, no switches at the command line. Have you downloaded a new version recently? Bjorn makes adjustments every time a new data set comes out that I can't open with IMG2PNG I'm pretty sure he modified it when the Steins data came out.

Posted by: djellison Jan 24 2012, 12:51 AM

I'll have another hack it it tonight, see what happens.

Posted by: Bjorn Jonsson Jan 24 2012, 01:25 AM

I haven't modified IMG2PNG for Rosetta yet - actually I was waiting for the Mars flyby data to appear. The problem is probably that when converting the floating point (level 3) images, IMG2PNG 'knows' nothing about the data so it uses the actual data range to map the floating point values to 0-65535 (the mininum floating point value is usually set to ~0 though). This is bad if there are bright specks in the images and makes them dark. Another problem is that different filters (or more accurate - images) get stretched differently. I need to upgrade IMG2PNG to 'know' the Osiris data. I may do it this weekend or possibly earlier.

Interestingly, the Messenger data is somewhat comparable to the Osiris level 3 data since it contains floating point data. The important difference is that IMG2PNG should know how to handle the Messenger stuff.

And by the way this is clearly a beautiful dataset - I get the impression that these are the best global views ever of Mars, at least if you only count framing cameras.

Posted by: djellison Jan 24 2012, 01:32 AM

AHHhh - I was using Level 3. I'll try again with l2 tonight.

D

Posted by: elakdawalla Jan 24 2012, 01:37 AM

Now you tell me I noticed the varying color values when attempting to assemble the 4-frame mosaic. I was hoping to make one with Phobos appearing four times across the disk, but had to give up partway through; it just wasn't working.

In the meantime, here's one with Mars limb and Phobos. Pseudocolor, made of Red and UV frames only (the reason for the bizarre violet color at the limb). I had to brighten Phobos a lot to make it visible at lower resolution -- it's a very dark moon!

For those of you attempting color processing, please note that the "Red" filter is actually pretty far red, out at 744 nm. For RGB purposes it's best to use the "orange" filter at 650 nm. So you want images whose filenames end in 82, 83, and 84 (orange, green, and blue filters in combination with a neutral density clear filter) or 22, 23, and 24 (orange, green, and blue filters in combination with a regular clear filter).

Posted by: ugordan Jan 24 2012, 06:25 PM

Why didn't you use the regular OGB filter set?

Posted by: elakdawalla Jan 24 2012, 11:56 PM

Because it was getting late, and somehow I missed that they were there. Ooops.

Posted by: Bjorn Jonsson Jan 25 2012, 12:36 AM

I have uploaded a new version of IMG2PNG. It has a new command line parameter "-fstretch" which is used to specify how to map the floating point data to 2 byte integers in cases where IMG2PNG doesn't 'know' which parameters are applicable (in these cases it has always used the actual range of the data as described earlier).

For example:

img2png *.img -fstretch0,0.1

This is useful for forcing IMG2PNG to use the same parameters for many images which is necessary when doing color composites. To get a rough idea of which fstretch values might be good first run IMG2PNG without -fstretch and notice a line in the output that looks like this:

Floating point to 16 bit integer conversion: [-0.000023,0.059897] -> [0,65535]

This suggests that -fstretch0,0.06 might be good - or maybe -fstretch0,0.1 to avoid getting saturated areas for images where different and 'brighter' filters might have been used.

The new version is available http://www.mmedia.is/bjj/utils/img2png/img2png_exe.zip.

(By the way, when you run this new version of IMG2PNG the first line you see should read "IMG2PNG v2.02 20120125" - otherwise you have an old version. There were some weird problems a few days ago when uploading and/or downloading IMG2PNG so you should check this just in case).

Posted by: elakdawalla Jan 25 2012, 12:56 AM

Bjorn, I just want to say here publicly how grateful I am for your continuous support of IMG2PNG. I know that you use the software yourself, which may be why you update it so regularly, but by making it available and useable by the public and maintaining it so up-to-date you are doing an immesurable service to our community. Thank you!

Posted by: ugordan Jan 25 2012, 08:29 AM

I had a feeling that might have been the case. Also happens to me sometimes.

Posted by: machi Jan 25 2012, 10:19 AM

The same is true for me. Different ways how open raw pds images exists (they can be opened simply as raw file in appropriate editor), but these batch converting capabilities of IMG2PNG are unbeatable. Thank you!

Posted by: Phil Stooke Jan 25 2012, 05:33 PM

This is the result of enlarging (x4) and merging twelve of the Phobos occultation frames. The noise partly cancels out leaving a clearer view of Phobos itself.

Phil

Posted by: Bjorn Jonsson Jan 26 2012, 12:38 AM

Thanks. In a sense this is a positive feedback loop: The better IMG2PNG gets the more planetary images I get to see - which is a good thing .

Posted by: Paolo Oct 1 2012, 11:27 AM

many years later... ESA has released one Rosetta Mars image as its image of the week
http://www.esa.int/esaSC/SEMAR9ERI7H_index_0.html

Posted by: ugordan Oct 1 2012, 12:16 PM

And it's Emily's composite at that.

Posted by: TheAnt Oct 1 2012, 02:47 PM

Yep, last in the credit E. Lakdawalla - small world this.

Posted by: bobik Feb 1 2013, 01:49 PM

In her http://www.planetary.org/blogs/emily-lakdawalla/2013/01311445-rosetta-philae-mars.html, Emily Lakdawalla reminds us of the famous Rosetta/Philae Mars flyby photo. But the caption is erroneous. The view is not centered near Syrtis Major. The photo shows the border between Acidalia Planitia and Arabia Terra. On the left side is visible the southern part of Cydonia Mensae, and on the right side is visible Mawrth Vallis and the Oyama crater.

Posted by: Phil Stooke Feb 1 2013, 04:44 PM

Here's a comparison to show that is correct.

Phil

Posted by: machi Feb 1 2013, 05:42 PM

Quick comparison with map:

Posted by: elakdawalla Feb 1 2013, 06:21 PM

Thanks, I've fixed the caption. Also, it's been pointed out in the comments to my blog post that CIVA-P is not a color camera; the photo was colorized before release. Shame on ESA for not mentioning that.

Posted by: machi Feb 1 2013, 06:39 PM

It's interesting that CIVA-P isn't color camera, so for color panorama, we'll need color images from ROSETTA's OSIRIS cameras.
Another Philae imagers CIVA-M and ROLIS are capable of color photography, but only in case of small area under lander (ROLIS) or even smaller (~3 mm diameter) extreme close-ups (CIVA-M).

Posted by: Explorer1 Feb 1 2013, 06:46 PM

Less than a year until Rosetta's hibernation ends, too (January 20th, 2014).

Posted by: stone Feb 2 2013, 11:38 AM

CIVA-P is only one of a whole bunch and it will make a panoramic image of the Philae landing area. The fact that additionaly CIVA-P and not only Osiris made images of Mars during swing by was that by luck they were able to ave it in their view. The Lander is stored at the outside of Rosetta andall the instruments are made to operate on the surface of the comet and not during cruise and fly bys.