This has been bugging me for a while (and is what prompted me to finally join this forum).

Is there a reason why JPL (et. al) smooth off the limbs of planetary bodies in images before they release them? Here's an example with Mimas (CICLOPS.org):

http://ciclops.org//view_media.php?id=6895

Compare it with a similar image (probably from the same encounter):

http://img75.imageshack.us/my.php?image=mosaic29ff.jpg

I like my Mimas with mountains, thank you very much.

It has to do with geometric reprojection of mosaic data. Cassini is somewhat notorious for the slow speed of taking images (typically on the order of a minute per frame for normal resolution images) and also the lack of a scan platform means the entire s/c has to slew to change the mosaic footprint. Both of those things take time and the spacecraft and the moon are moving (very rapidly in Mimas' case, it's low in Saturn's gravity well) and the perspective changes rapidly. Just mosaicking the footprints together causes problems with image seams due to the perspective shifts. Geometrically correct mosaics such as that one take into account the perspective and reproject actual image data on a shape model of the moon (say a triaxial ellipsoid at most) . This has a nasty sideeffect of giving a flat limb since the model doesn't have topography and some moons are very rough, but the end mosaic is scientifically more accurate. Fudging the mosaic footprints gives aesthetically more pleasing results, but they should not be regarded as scientifically accurate or used for serious mapping for example.

Those Mimas images are mosaics? I thought they were single frame images.

Besides, I figured the CICLOPS.org page was more aesthetic than scientific.

Gladstoner - I'm with you - Give me mountains on Mimas. Can't wait for the extended mission sub 10,000km flyby
I think Gordan is right - the ciclops image you link to IS a mosaic, but only of two frames - shuttered in August 2005.
My guess would be that it was easier for whoever spliced them, to assign a symmetrical limb during processing, so that both frames of the mosaic could be brightness-matched satisfactorily and quickly.
It's aesthetic. But wonderfully, we all have the option of referencing the PDS (or the NASA raw image page) for the unabridged data whenever we wish.
Once again for the record - fabulous outreach, congratulations to Linda S. and the team.

That mosaic has four-frames actually. The speed of the flyby prevents just "adjusting" each frame so they fit. The best way fitting the frames together is to map project each frame into an orthographic map projection. The unfortunate side effect is that the limb is smoothed since the model used doesn't include all those little bumps on the limb. While annoying, it is the nature of the beast and it is better than the alternative in my opinion.

However, in my last few mosaics, I have combined two mosaics: my usual map projected mosaics and limb mosaics that use that original frames. All the seamlessness of the map projected versions with all of the limb-y goodness.

All this is missing the point - the ultimate goal of the satellite imaging is to create global high resolution photomosaics of each moon. Regardless of the issues in any one flyby sequence, eventually we need to combine numerous images to make a global mosaic - such as Steve Albers has been doing - and then it can be projected onto a sphere for display from any orientation. Most Earth, Moon or Mars globes don't include relief (of course I know it wouldn't show up as much as on Mimas). There might be a chance later on for a high resolution global DEM to be used to create a more realistic global visualization complete with limb photography. But that's a long way off.

We cartographers don't like relief anyway - it distorts things. I won't be happy until all the craters have been filled in and all the mountains bulldozed, so Mimas is like a billiard ball. And it will be easier to map, too.

Phil

Phil, do I sense some exasperation relating to that pesky potato-shaped rock Eros you're working on??

Blast all potatoes!!!

Phil

Yes, but considering it doesn't have much in the way of albedo features, once you bulldoze it, who will want a map of the place?

Bulldozing contractor to write up the invoice

Ted said "Yes, but considering it doesn't have much in the way of albedo features, once you bulldoze it, who will want a map of the place?"

That's a problem for the marketing department. I just make the maps. And no map could be more precise than a blank map of a featureless object. Besides, think of the money I'll save on printer ink.

Phil

"Blast all potatoes!!!"

Oh, Stop spuddering.

Incidentally, a printer could run out of magenta, yellow, and black ink, and you could still print an accurate map of Uranus (at Solstice at least).

On a side note....

One day, my inkjet printer decided to stop printing in cyan, and all color print came out as various shades of red and pink. Considering that much of my print is finder charts for comet observing, red ink can be quite problematic when using a red flashlight outside at night. Actually, it's quite useless. I had to pencil in some of the field stars at the time, and later deal with some wierd texture with black ink until I got a new printer (b/w laser, thank you very much).