Originally posted by ArchAngel
"Which 36-bit value would uniquely identify any form of chlorophyll? "
There is none. It would depend on the brightness.
I am not only concerned with chlorophyll. It was one example to begin with.
It goes beyond just the brightness, and you know it. You simply can't tell WHY the pixel in question looks green with basic RGB filters... you just
know that it
does.
If RGB can't tell the difference between life and a non-living background, then RGB is life-blind.
The cam is color blind.
Yawn. Every CCD camera is color blind. So are humans. Who cares? It's a non-point, so long as you choose to misinterpret what the data actually
reveals, instead of what you think it NEEDS to reveal.
Many years ago I was forced to edure an episode of "The Benny Hill Show'. In one skit two people were playing ping-pong. It was normal until one
point where the ball began jumping around in an impossible way. Was it a digital trick? Not on their budget. There was a man dressed in a black outfit
against a black background with a rod that had a ping-pong ball attatched to the tip. The man was invisible because he matched the background in all
spectrums that the camera used in its RGB format.
In a similar way things can be hidden against the background. Just like the JPL symbol disappears in some filters, and not in others.
Yes, the JPL symbol "disappears" (as you call) it on that filter, just like it "disappears" on the red filter of both of the filter sets that you
pointed to initially.
Gasp! The RGB filters can be selectively used to make things disappear!
It's actually much, much worse than that. With the RGB filters, it's not hard to come up with scenarios where actual chlorophyll-B laden life
scattered over something like an olivine base would "disappear" on EVERY SINGLE RGB IMAGE FRAME, as well as the combined output. No matter what you
select, it's always "gone".
There are an infinite number of different response curves that will give IDENTICAL results for each of the R, G, and B filters, precisely because they
take an entire "area under the curve" for such a (relatively) wide range of frequencies. You could have an infinite number of combinations that
blend perfectly into the background, despite the fact that they are actually different in their individual frequency responses... in a huge number of
those cases, radically different responses.
It would be the perfect analog to your "Benny Hill" example. By the standard you have set, we must therefore conclude that RGB filter sets are life
blind!
Q.E.D.
Wait, it's even more insidious than I imagined! Human eyes also use a process analogous to RGB camera filtering (three overlapping "totaling"
light curves), and the algae in question would disappear for them, too! Oh my God, the alien overlords have engineered our race to be life blind!
The Pancam filters can detect a wealth of information that is utterly hidden from human eyes, and use it to accurately isolate different components
that are "invisible against the background" to the human eye.
Simple RGB filters cannot. The multispectral nature of the Pancam is therefore superior for the purposes of IDENTIFYING what is actually present,
instead of just making pretty pictures that match what our obviously alien-engineered-life-blind eyes see.