Good evening Alex...

The question of "the" MSSS destreaking algorithm's gotten pretty muddied, and it would help if I put it more precisely: I am aware of one destreaking algorithm implemented by MSSS in C that looks at the column-to-column differential and preferentially boosts the coefficient; it responds to the slope of the data inside a variable-sized sliding window. I got a copy from, I believe, the USGS FTP software archive. They've no doubt tried out many other algorithms, so it isn't correct to call this the MSSS algorithm. Sorry about the confusion.

MSSS applies no explicit filtration of any kind to the image archive. Obviously, the raw data is raw and has not been processed in any way.

Of course not. We were talking about using the MOC-O-Matic on raw data.

If one doubts MSSS' scientific integrity there, feel free to go back to the raw images.

Your protectiveness toward MSSS is commendable, but I assure you, once again, that it's not necessary. There was nothing in my message that could be taken as impugning the scientific integrity of MSSS. Relax, I have no beef with Doc Malin.

Those "martian spiders"...

True, though in all honesty many of the features Greg points to bear little or no resemblance to the "spiders" that Kieffer presented at the Reykjavik conference, which I had the good fortune to attend. The "gullies" depicted in MOC NA M09-05128...

Yes. I was ranging a bit afield there and thinking of the theme of "martianspiders.com" in total, which has some examples of what, I believe, Kiefer's talking about. Including that original image of the radial features that got AC Clarke buzzing about gigantic trees...I do find Kiefer's idea much more plausible!

I appreciate your thoroughness in going back through the threads and getting caught up. Always an interesting conversation--thanks.

--Robert