Printable Version of Topic

Posted by: djellison May 9 2008, 09:11 PM

Finally - I found an image of the Phoenix SSI with, I presume, two spare MER CCD's inside

http://www.lpi.usra.edu/meetings/lpsc2008/pdf/2156.pdf

It's sprouted a big lump on top of the white cylinder MPF and MPL design, due, I presume, to the use of two CCDs instead of one.

Doug

Posted by: Deimos May 23 2008, 03:17 AM

Additional information is online.

http://www.met.tamu.edu/mars/

Mark

Posted by: elakdawalla May 23 2008, 03:51 AM

Oooohh...juicy info that only a true space enthusiast could love...

http://www.met.tamu.edu/mars/filenames.html
Note that there's an indicator in the filename for whether the image is part of a mosaic or stereo observation (or both). Coool!

Et cetera...

And how the heck did I miss the fact that you were the lead for SSI, Mark?

Is there anything you can tell us about the nominal plan for the pipeline for images? Peter Smith said at today's press briefing that "they will be made available on the Internet as they come down. You'll see them as I see them. We plan to operate the mission in an open manner." Which I had heard was true but it was very nice to hear him say that in the official context of the press briefing.

Can you hear that? That's the sound of me rubbing my hands and licking my lips -- I can't wait to see new data from a new spacecraft!

--Emily

Posted by: climber May 23 2008, 04:18 AM

From Mark links, I hope to this very soon (with another background):

Posted by: djellison May 23 2008, 07:40 AM

It's also the sound my me click the PTGui icon. Rapidly.

doug

Posted by: mhoward May 23 2008, 01:12 PM

Also, in the downloadable version of the Phoenix clock, the index.html file gives the source code for converting to "Phoenix" time. Cool.

One question on the Phoenix filenames information: Is the spacecraft clock time interpreted the same way as MER's? Ie, "number of seconds since January 1, 2000 at 11:58:55.816 UTC".

Also, I could ask if there are any plans to release metadata, especially camera pointing information, for the images... but I don't want to bother Mark as we get into suspense time here.

Posted by: slinted May 25 2008, 11:13 AM

Here's a rough graphic showing the hue through the SSI filters in visible light using the center wavelength and FWHM from http://www.met.tamu.edu/mars/SSI_filter.html. I've excluded the solar filters, the two diopters, and all the infrared filters to focus on the filters that can be used for color images of the surrounding terrain.

http://www.lyle.org/~markoff/phoenix/ssi_visible_filters.jpg
http://www.lyle.org/~markoff/phoenix/ssi_visible_filters.jpg

Posted by: mhoward May 28 2008, 02:52 AM

I just found out something remarkable via #space. The 'large' size images on http://fawkes3.lpl.arizona.edu/images.php?gID=0&cID=8 have metadata tags embedded in the JPG. Just open the JPG file up in a text editor to see.

Posted by: elakdawalla May 28 2008, 03:09 AM

Remarkable...yes...

OK, mhoward, why haven't you unveiled the Midnight Phoenix Browser yet?

--Emily

Posted by: slinted May 28 2008, 03:10 AM

Sorry I didn't point this out earlier. I thought it was mentioned in yesterdays main thread. It's a meta-data dream come true! The large version images on the LPL site each have a caption and a series of tags embedded in the "Comments" section of the JPEG headers. http://www.sno.phy.queensu.ca/~phil/exiftool/ might be a good tool for recovering them. The Python Imaging Library can also read them (specifically, the app["COM"] method of an Image object).

Here's what the tags look like for a raw image (http://fawkes3.lpl.arizona.edu/images/gallery/lg_532.jpg):

This image was acquired at the Phoenix landing site on day XX_DAY_NUMBER_XX of the mission on the surface of Mars, or Sol 1, after the May 25, 2008, landing. The SURFACE STEREO IMAGER RIGHT acquired this image at 14:49:43 local solar time. The camera pointing was elevation -25.8395 degrees and azimuth 280.315 degrees.

PRODUCT_ID = "SS001EFF896308546_10D10RAM1"
FRAME_ID = "RIGHT"
FRAME_TYPE = "MONO"
INSTRUMENT_NAME = "SURFACE STEREO IMAGER RIGHT"
LOCAL_TRUE_SOLAR_TIME = "14:49:43"
RELEASE_ID = "0001"
SOLAR_LONGITUDE = 77.1473
FILTER_NAME = "SSI_R10_604NM"
EXPOSURE_DURATION = 408.0
INSTRUMENT_AZIMUTH = 280.315
INSTRUMENT_ELEVATION = -25.8395
PLANET_DAY_NUMBER = 1

Posted by: ahecht May 28 2008, 05:52 AM

I used that page to create a Google Gadget so that you can have the Phoenix Mars Local Mean Time on your Google homepage:
http://www.google.com/ig/adde?hl=en&moduleurl=hosting.gmodules.com/ig/gadgets/file/116148110345150166419/phxtime.xml

Posted by: algorimancer May 28 2008, 12:28 PM

Seems to be from 1980.0 (UTC).

In APG I handle the conversion from the system clock to Sol & Local time something like this:

#define SECSPERSOL 88775.244
#define PHX_SOL0 896166627//seconds since 1980.0
unsigned int uiSysClock,uiSol,uiH,uiM,uiS;
uiSol=(unsigned int) ((uiSysClock-PHX_SOL0)/SECSPERSOL);
uiSysClock=uiSysClock-(((unsigned int)uiSol*SECSPERSOL)+PHX_SOL0);
uiH=uiSysClock/3600;
uiSysClock=uiSysClock%3600;
uiM=uiSysClock/60;
uiS=uiSysClock%60;

I just figured this out from a variety of sources (and yes, the code could be a tad better optimized), but it seems to match the image products. One difference from MER is that the first Sol is Sol 0 for Phoenix as opposed to Sol 1.

Posted by: algorimancer May 28 2008, 12:47 PM

That is indeed interesting... now if we just had some handy utility to batch download the images and correct the file names.

Posted by: pechisbeque May 28 2008, 02:07 PM

That's really nice! Is there also one for the Rovers?

Posted by: jamescanvin May 28 2008, 02:22 PM

Well I kind of have something that does that, however it is in no way ready for general use (particularly as most of you will be using Windows). However I have chucked the results on the web if it is any help to anyone:

http://www.nivnac.co.uk/phoenix/raws/

I'll try and keep it up to date. I've included a .zip of each sol to save some effort downloading.

Now that's done, time to get my MER panorama software to recognize Phoenix data.

James

Posted by: ahecht May 28 2008, 03:02 PM

Yes, but I can't take credit for it -- simonbp created it:
http://www.google.com/ig/adde?hl=en&moduleurl=hosting.gmodules.com/ig/gadgets/file/108711613412433434334/MarsSundial.xml

Posted by: AndyG May 28 2008, 03:03 PM

...Shouldn't that be the Midnight-Sun Phoenix Browser?

Andy

Posted by: pechisbeque May 28 2008, 03:27 PM

Yours is cooler because it has the Sol!

The times for Phoenix between the two gadgets are off by 2,5 minutes... Are you taking relativistic effects into account in your clock or what???

Posted by: Del Palmer May 28 2008, 04:16 PM

Forget Rob Manning -- you're my new hero! The Zip file is a huge time-saver; constantly right-clicking to save hundreds of images was getting a little old...

Posted by: elakdawalla May 28 2008, 04:21 PM

You are a god, James.

--Emily

Posted by: Deimos May 28 2008, 04:50 PM

I put the one with the sol on my iGoogle page (thanks!). It matches mars time on my computer to within ~1 sec (using my own system clock I assume). That matches time displayed in the SOC to <15 sec. Mine's right ;^) Unless you believe Einstein, then there's a 15 minute light-time ambiguity. That ambiguity is really annoying ... so I choose not to believe in Einstein.

Posted by: algorimancer May 28 2008, 06:29 PM

That's wonderful -- just what I was looking for

Posted by: algorimancer May 28 2008, 06:39 PM

Mark/Deimos,

Enjoyed your talk on today's press conference, however I rapidly became lost keeping track of the various reservations for the robotic arm workspace and wondered whether you might be able to put together a map overlay highlighting the trough between the polygons and the various designated preserves/workspaces.

Posted by: jamescanvin Jun 1 2008, 02:46 PM

Can someone explain to me why some of the SSI images taken through different filters don't have the same pointing? The difference doesn't seem to be reflected in the metadata info imbedded in the images either, so I'm finding it all but impossible to correct for it automatically in my processing software.

Cheers,

James

Posted by: ugordan Jun 1 2008, 02:58 PM

I wondered about the same thing, James. Could it be due to filter optical characteristics? I'm reminded of a very similar thing with Cassini's WAC, the RGB filters are consistently offset up to 3 pixels from each other's boresight.

Posted by: Airbag Jun 1 2008, 03:04 PM

Perhaps the wind is moving the SSI slightly in between images?

Airbag

Posted by: jamescanvin Jun 1 2008, 03:16 PM

I don't think it can be optical characteristics as some images have the shift and some don't, it is very inconsistent as far as I can see.

EDIT: On further investigation it looks like the shift between the 1,B and C filters in the 3 problem frames in the sol 6 mosaic I posted in the other thread is approximately the same in each of them. But why are the other 5 fine? I'm confused!

Posted by: Deimos Jun 2 2008, 12:13 AM

Working that. There is a zone where elevation backlash is acting unstable. I have an idea what to do, but in terms of getting new diagnostics up there are still bigger fish to fry, so it may be a little while.

Posted by: Tom Tamlyn Jun 2 2008, 03:00 AM

Informative http://phoenix.lpl.arizona.edu/blogsPost.php?bID=202 by Mark Lemmon
discussing why taking pictures with a robot on Mars isn't exactly point and shoot.

Mark, when you get the opportunity I'd be grateful for more detail on why and how
Phoenix imaging operations are different from working with the MERs.

Also liked the discussion of why the title was appropriate.

TTT

Posted by: jamescanvin Jun 2 2008, 07:18 AM

Thanks for the reply Mark, I guess I'll work on my own fix to realign the images in the meantime.

Posted by: Airbag Jun 6 2008, 05:14 PM

With all the attention on the filters, let's not forget about the boom that deployed the ISS:

http://www.aec-able.com/Booms/coilboom.html

Looks like they can add a new entry to their "successes" table!

I've always been fascinated by these "tensile" structures. Although it doesn't use quite the same principle as the coilABLE boom, the Hirshhorn museum in DC has a really great example of such a tensile structure; you can stand right below it and wonder how it stays up with just wires and rods (and the rods don't touch each other) and also how it was raised to begin with.

http://www.flickr.com/photos/cybertoad/2415466249/

Airbag

Posted by: Deimos Jun 9 2008, 10:40 PM

An attempt at browseable access to the raw images has been added to the SSI site. The format borrows heavily from Emily's, with some extra info. At the moment RAC and OM are not included. The page will be hours behind the UA image page, but it is hoped the organization appeals to a certain small corner of the net.
http://www.met.tamu.edu/mars/

Posted by: djellison Jun 9 2008, 10:44 PM

Mark - that's just ideal - perfect. Given how hard you must be working at the moment, this is above and beyond the call of duty!

Incidentally - I was hoping that the descriptions would help me figure out what one set of obs was....but I'm still none the wiser "11DE-0: SSI_MS_OFB_1200-1300" - is that part of the SSI hardware atop the mast or something? It looks a bit like some of the various brackets and so forth on the lander deck, but far too big.

Doug

Posted by: Shaka Jun 10 2008, 12:55 AM

Ya gotta love it!

Posted by: Airbag Jun 10 2008, 01:12 AM

Man, I almost hate to mention this...but the Sol 15 page appear to have what looks like an html bug ("v>") towards the end:

Perhaps related to there not being any images (yet?) on that page?

Airbag

Posted by: James Sorenson Jun 10 2008, 03:00 AM

Why isn't the "Peter Pan" Mission success panorama being done at full resolution color?. I would think that they would be doing this at full resolution. Or are they gonna do that later after this pan?.

Posted by: ugordan Jun 10 2008, 08:09 AM

I was wondering about that as well, James. Seems like a waste of bandwidth to first be doing a 2x2 binned panorama only to redo it again at full res at some point in the mission.

Posted by: djellison Jun 10 2008, 08:46 AM

MER did it.

Twice.

Posted by: fredk Jun 10 2008, 03:01 PM

It's great to see this page - thank you! And the "extra info" is certainly appreciated...

Posted by: Deimos Jun 10 2008, 07:45 PM

This is a really good question, and like all such, it has multiple answers.

One answer is about quality. On that front, the color pan from SSI is actually closer to full resolution than, say an image from your digital camera (making assumptions here..). The difference is your camera hides that fact, SSI is open. A typical Bayer pattern CCD takes images such that the green are half-resolution and the red and blue are quarter resolution. The image is manipulated on chip or in software to predictively extrapolate to full res before you ever see it. SSI has some filters at full res, others at quarter res, and they can be upconverted even better than the Bayer-sampled images can be (see the sol 2 pan press release, based on 1 full res and 2 1/16th res filters).

OK, the pan will be full res, but why not do better? Well,there's the practical answer. We need to see the pan before it is useful to us. Finishing it on sol 150 is not an option. So, design a pan that can be acquired and downlinked in a reasonable time. This pan is a requirement for full mission success, and a useful planning tool (interpretation of immediate workspace through context, identification of interesting distant features). So you figure out how many frames you want, how many sols you have to finish, and what fraction of the downlink you can use (unlike MER, the panoramic camera is merely one of 3.5 major data hogs). Then, you land, get great comm and thermal, end up with more data. So the runout version of the pan is obsolete, and the new and improved pan is in. Another practical issue is that Phoenix has very little flash memory. The pan is the last thing saved, almost, so we cannot just take lots o' data and play it back at leisure.

Isn't it a waste to take multiple pans? Well, no. An "efficient" way to acquire a pan would be to spend the whole mission getting the best quality. That would mean on sol 45 *half* the site is unseen. Thus we plan not 2, but 3 pans. The sols 1-3 site pan, low res, for quick look. The full res color pan to sol 30(??). Then, maybe, if needed, a high res, low compression, many filter pan. But if that never happens, we're still good. (Actually we need to reacquire the low res site pan a couple times, just as pancam does, as part of an orbiter-coordinated experiment.)

Posted by: vk3ukf Jun 11 2008, 11:20 AM

Oh, these pages are heavenly, really.

May I ask you folks if I can host any links to your own MER and Phoenix software on my own MER software page. I was also going to weave the Phoenix codes into my MER software as well.

http://www.vk3ukf.com/vk3ukf_files/MERDAT.htm

I need to read more here, I also need sleep badly.

Back in day or two for a look see.

VK3UKF.

Posted by: climber Jun 12 2008, 10:36 PM

From Mark row images pages Sol 17 : http://www.met.tamu.edu/mars/017.html
Is this a confirmation that Steve the Cat is really on bord ?

SSI 13F - CheshireCat 1248 15 9
SSI 13F - CheshireCat 1249 15 0

Posted by: Tman Jun 17 2008, 09:51 AM

Currently it isn't possible to match these two "Peter Pan" pictures regarding brightness of the hills, except doing extensive manipulations that, however, cause no satisfying result:

http://www.greuti.ch/phoenix/peterpan_az220_az231.jpg

It is probably due to the different time of sol (3hrs) that caused the different brightness, isn't it.
- left picture of that mosaic http://www.met.tamu.edu/mars/i/SS018EFF897813205_11E1ER1M1.jpg (13:42:53 )
- right picture http://www.met.tamu.edu/mars/i/SS017EFF897713309_11E1CR1M1.jpg (10:42:20 )