Wow!

Blimey, that's ineffably stupendous, Matthias!

Matthias,

I am amazed by what you have accomplished. (I just recently learned to stitch a series of photos into a simple flat panorama.)

But when you have mapping experts like Phil and image experts like Gordon extolling your work, you know you have really accomplished something. Congratulations on the success of your shape model.

David

That is astonishing.

It's impressive and beautiful, Matthias!

Thank you guys!

It has been a lot of work but it is finally starting to pay of.

I updated my shapemodel a little. Revrote the code and strange macros that I had to make to do the sillhouette space-carving that takes care of the areas where I do not have stereo coverage. I also found a small OSIRIS sequence that I could use for stereo in the esa landingsite conference movie. (compressed to crap but i got some data out of it by averaging a lot of frames)


I'm hoping to show you some more stuff in a day or so.

Some new NavCam images showing the plumes in spectacular detail.
CometWatch – 26 km on 26 September

Congratulations for your amazing 3D modeling and imaging work Matthias!

Boy that is not a very good composite of the 4 photos. You can see clear match points between the photos that are widely separated. Were the photos taken from vastly different spots in orbit so the perspectives didn't match well?

As the orbit is lowered, the orbital velocity increases, so images taken at equal time intervals will have more offset than those taken in a higher orbit.

because it is not a composite .

It is just the 4 side by side

Quoting the article:

Challenge accepted:)

Assembled and made into anaglyph stereo.
both as a still and a movie:
http://mattias.malmer.nu/2014/10/cascade-of-light/

Majestic... many thanks to Malmer for crafting & sharing these images and animations that bring 67P to life...

The two movies you've posted are just incredible. Chelsey Bonestell's jagged little worlds finally realized. Guess we've been dreaming about this for 60 or 70 years.

Several preliminary data have been made available, including a preliminary 3d-shape model.

That's an awfully low density--0.4 g/cm^3. Must be a significant void fraction, I'd guess. Are there any really low density phases of water ice or more exotic ices?

It's not a solid ball of ice, there are a host of other materials in there, including dust. In fact one slight worry is the lander could sink into meters of dust, which would pretty much be the end of the lander mission.

Malmer, thank you so much. I'd love to share your images on Facebook with friends but I just know many of them do not have red/blue 3D glasses. Could you produce one more image that's a non-anaglyph please?

If you look at my blog post featuring Mattias' images, you'll find I posted all his stereo pairs as 3D anaglyphs, and left-right pairs, and several as flickr gifs. There's even a link to a zipfile containing everything!

I will rerender the animations in a more user friendly way. Youtube have got 3d support so im going to experiment a little with that. You can choose from a bunch of different display modes. 3d and non 3d. I will post it here when I get it right.

Malmer,
in view of your fabulous comet visualisations I feel it would be most appropriate when ESA sends you a VIP invitation to attend the Philea landing event in Darmstadt!

It looks like the jets are impinging off the upper right lobe of the comet (the "chin" of the head). Or is that just an effect of the changing viewpoints during the capture of the four images?

Managed to get Microsoft ICE to assemble the mosaic. I think it came out pretty good. Image further processed to remove noise and improve constrast.

Click to view attachment

Full-sized: http://l3.picsurge.com/o9KH6W/ESA_Rosetta_...926_stitch2.jpg

That's very nice. And look - a new image set from only 18 km:

http://blogs.esa.int/rosetta/2014/10/06/co...e-j-from-18-km/

The closer we get, the rougher it looks.

Phil

Very handsome mention of Matthias' work on that ESA blog page.

The overlap of A and D1 allowed for an anaglyph:
Click to view attachment

I absolutely agree. I just cropped out images of a boulder seen twice the 14 and 30 September (2.5 m/pixel and 1.4 m/pixel).

Click to view attachment
The boulder seems to be made of rounded blocks ~25 m wide (white arrows) reminiscent of conglomerates seen on Itokawa.
Are the strata of the comet composed of conglomerates? Did the rounded blocks accreted on the comet parent body? Fascinating object!

ICE stitch: http://l2.picsurge.com/8rzoMg/ESA_Rosetta_...0_D1_stitch.jpg

I have reprojected the limb of the 'head' from the Sept. 30 image as if it is all one long straight ridge. I never expected such dramatic scenery.

Phil

Click to view attachment

Todays Composite and Anaglyph:

http://mattias.malmer.nu/2014/10/comet-on-september-30/

Cool of the ESA blog to mention me like that. Bit proud of that

I'm struggling with delivery formats. How should I post the stereo animations? Anaglyph seems very 80:s tech. But youtube have managed to mess up displaying of 3d tv.
It used to be possible to just choose from all these options when watching 3d content. But now that little options menu is gone (or I cant get it to work at least on my machines)
I could post side by side stereo that would work if plugged into a 3d TV like this: https://www.youtube.com/watch?v=OCYTcW3ctjU. But it looks like crapola for anyone else.

I could perhaps post a fallback mono version. But then you would have to know how to find them on youtube. (my youtube fu is not very strong yet so I'm not sure on how to link them together in a coherent way.)
What do you think?

Probably the best way is to let people download two separate movies - a left and right eye - and then people can use something like Stereoscopic player ( http://www.3dtv.at/Index_en.aspx ) to play it in any format they choose - Anaglyph - X-Eye - Over and Under etc etc. If hosting them's a problem, I'm sure UMSF and/or TPS can help

My Fuji 3D camera produces 3D videos in .avi format -- when I play one of these, Windows Media Player opens two synchronized video windows playing left and right eye, and I can position them side by side to view it cross-eyed or wall-eyed. Can you output in that format? It's not as good as streaming it on Youtube. I wonder if there's a way to have two Youtube windows in side-by-side players and make them play simultaneously?

(Example output from my 3D camera here, 72 MB)

-- for the NON Microsoft Windows crowd --

VLC 2.15 also will open two windows for the video
https://planetary.s3.amazonaws.com/emily/20110404-115900.avi

As far as still images, I'd vote for including the separate L and R frames, in addition to the anaglyphs. For studying fine detail, nothing beats cross-eyed (or parallel) of two ghost-free images. Of course you can always extract L/R frames from an anaglyph, but if it's in jpeg format, some residual ghosting will always be there (and most obvious in the extreme contrast of a subject like C-G). An alternative would be an uncompressed anaglyph (eg png) from which it should be possible to extract the original L-R frames without ghosting.

For side by side youtube stereo you could use http://www.youtubedoubler.com
(Bit of a hack solution...)

Thank you all for the input... lots of things to try.

I'm going to experiment a little with youtube because it is supposed to be able to display 3d content in whatever way you want to. but it is something a bit strange with the combination of html5 and nvidia 3dvision going on. (it used to work back in the day)

I want to have something that everyone can easily watch. People on this forum would go trough great lengths to view a space image but I would also like to target the more casual observer...

For stills I will post left, right and anaglyph versions. Perhaps even a mono version aswell (the warping to create the left or right can sometimes introduce unwanted artifacts)

M

My take on the mosaic of 30 September 2014:

I agree that it is very important to target the casual viewer: outreach is of paramount importance.

For easy viewing I think Youtube is a good solution. I would suggest two versions:
- a red/green (for those with glasses from an astronomy mag)
- a cross-eye side-by-side, something like this: https://www.youtube.com/watch?v=zBa-bCxsZDk&gl=BE (the dot really helps cross-eye novices)

http://blogs.esa.int/rosetta/2014/10/08/co...us-on-the-neck/



Fantastic new image - and for the first time I have seen an unmistakeable track left by a rolling rock, right there in the neck.

Phil

Do you think that's a rolling boulder track? Or is it a fissure with collapse pits?

Amazing shot. These boulders - I understand that the photo scale is 1.45 meters/pixel but it's really hard to determine from just looking on the internet shots without photo editing software that lets you zoom way in, just how big a pixel is. Could someone give us an idea just how big across these boulders are and, given the gravity that's been calculated, what would they weigh? I would imagine they are massive compared to the size of a human but could they be rolled around or lifted if the person had secured footing. That concept is just unimaginable to me.

A qualitative model calculation:

Take a boulder the shape of a cube with an edge length a little less than 14 pixels. Assume a density of 0.4 kg / liter.
The volume would be (20 m)³ = 8000 m³. The mass would be 8000 m³ x 400 kg / m³ = 3.2e6 kg ( = 3200 metric tons).
With a mass of 1e13 kg for the comet, the surface gravity at 2 km distance of the barycenter (neglecting rotation) is GM/r² = (6.672e-11 Nm²/kg² x 1e13 kg) / (2000 m)² = 1.668e-4 N / kg, hence 3.2e6 kg x 1.668e-4 N / kg = 533.76 N. That's about the weight of 54.4 kg on Earth.

The actual weight of a boulder depends on its size, shape, density, distance from the barycenter, distance from the rotation axis, details of the field of gravity of the comet. But the numbers should provide an idea of the order of magnitude.

Edit: Here a scale (right of the center):
Click to view attachment

For what it's worth this is exactly the place (under the 'chin' of the head portion) where I would most have expected to see boulder tracks or landslips. It's the place where the surface is most steeply sloping relative to the gravitational equipotentials. See Y Bar Ranch's simplified gravity model here: https://lh3.googleusercontent.com/-5-90J06v...simpleComet.jpg It would be amazing to watch the ultra slow motion roll of one of these boulders! Perhaps someone can work with Gerald's acceleration value to produce a plausible real time animation of the process.

On a 30° slope it would take at least 187 seconds for the first pixel, 264 s / 2 pixels, 374 s / 4 pixels. I've neglected friction and rotational momentum, slowing things down even further.

I expect you could roll or even lift one, but you'd have to take your time. They may only weigh a few kilograms but they're likely many tonnes in mass. So: keep pushing hard for a minute or so and maybe you'd notice some movement starting. (I did this experiment once with a ship at the dockside.) Also, once you've got it moving take care to stay out of the way!

My take on the 2 october mosaic:


And a version with levels adjusted to highlight the jet:

The direction of the track lines up with the crack we have seen in earlier navcams on the side of the neck. I think we are seeing material that have sunk down into the crack. It looks a bit like sand in a hourglass...

Beautiful!

And a quarter-width left of centre, a clear crater.

I've met people who watch movies like this. They're moth enthusiasts. It can take hours for the moth to wake up sometimes.



So, maybe it's not a boulder track but a crack. It's equally interesting either way so I have no preference. Nevertheless I note two things. 1/ There are boulders at the bottom of the slope. Presumably they got there by rolling (slowly) down. 2/ We recently saw an excellent example of what appeared to be a crack in the neck. There was no obvious sign of material collapsing into it.