Here's my false-color composite of the first target on Meridani plains - Named "homeplate" by JPL. In the large panorama it appears to be bedrock exposed by the wind:
http://marsrovers.jpl.nasa.gov/gallery/pre...360_cyl_br2.jpg

But in the newest Pancam closeups (subframe), it looks strange... At first I thought it was a smudge on the Pancam lens - but it macthes the position in the Nacam footage.

If I didn't know better I'd say it almost looks like muddy water.
(But I'm sure that's just the low detail and compression)

See my attached composite (L456 adjusted false-color):
(source: http://marsrovers.jpl.nasa.gov/gallery/all...unity_p058.html)

could it be very fine dust? The "far side" of the dunes have about the same colour, so maybe it's the angle of looking that makes it look like that

are there 3-D pictures available of these pictures?

They do have variable compression, which it looks like they ramped up pretty high for this series. The raw files might also have been missing some small portions of the downlink data, which in ICER makes for less detailed, messy images (and if its bad enough, the black boxes). If thats the case, we might see more detail if they re-downlink them with better success.