We also found that there was an over-aggressive snow removal algo from the existing color data, that was making desert colors quite bland.
Funny how I never realized there was something missing in desert areas until you shaw this new picture 
I did. You can see from some earlier Outerra pictures taken from orbit that certain surface features were much more prevalent..... then they were gone, and I assumed the culprit was atmospheric scattering..... Which I started fiddling with.
This occasioned some discussion regarding unrealistic expectations caused by enhanced earth coloring in pictures.
Whatever the cause, I'm really glad to have the additional clarity back! (Yay!)
In fact. Just i didnt remember the times where the detail was there.
Just an idea about snow filter. Would not be posible to apply with some kind of mask which take in account the altitude and latitude? I mean. In the middle of africa is pointless to apply at a sea level. Maybe we can find snow there over 5.000 ft or so in so little points (Atlas, etc).
In Europe, depending on the pictures data, in summer will not find snow below 3.000 ft. And so on.
I dont know how would be posible to make those kind of masks, maybe with inteligent filters which make predictions where can really be snow by latitude/altitude or so (while altitude in a 2d pic is a dare...). But this would avoid the washing effect in the entire world and prevent the pure color in many % of the areas.
Anyway it seems this less agressive way respect better the colors
PD: a tricky way wouyld be take a 2d pic showing altitude by colors or so, and make the picture as big as the 2d map texture, vectorize the part in the altitude target, and apply the filter inside vectors.
