Duck Bay, Victoria arrival point Options

[Shaka]

Maybe it's some kind of martian "Mt Rushmore", eh?

Either that or a crowd of Rock People gathered together to gawk at the Alien Earth Creature.

"...Move along, Folks, it's nothing but a government experimental vehicle..."

[Jeff7]

Finally some colour filters to make nice pix...!
And hey, Hoagland... you want "Faces on Mars"? This place should keep you busy in blog posts for years. Maybe it's some kind of martian "Mt Rushmore", eh?

I see Elvis!!!!

[djellison]

L7 typically has the best spacial resolution.

Doug

[CosmicRocker]

Now I am really confused. I always thought the short wavelength filters seemed to have the best resolution, but when I said so sometime last year, some of the optics experts came out of the woodwork and chewed my butt off. Can we really notice more precise angular separation between the long and short filters? I seem to remember something about the cameras "not being diffraction limited at these wavelengths."

[Bill Harris]

You are correct. With diffraction limited optics the resolution is wavelength dependent. But the Rover optics are not "perfect" so the theory doesn't apply. I don't understand why, but the L2 (infrared) images appear grainy and noisy while the L7 (violet) images appear sharper.

--Bill

[Nix]

The iron-rich dust has a distinctive red hue and is imaged directly through L2,3..so images through those filters should be 'clearer' at times when there is less dust in the atmosphere and 'grainier' when there is a lot of dust around, scattering red light. Also important is the size of the dust particles vs the wavelength used to image them..

They used L7 for this reason to image the Gusev walls around sol 85.

I hope I'm putting this into words correctly

Nico

[helvick]

Nico,

I think you're right in broad terms but I don't think that any of that can account for the "fuzziness" of most of the images that we see - we see "fuzzy" images of items that are close up as well as far away with the amount of dust in the atmosphere has rarely been high enough to cause any visible effect near (ie within say 50m of the rover). We're talking about Tau value's that degrade light intensities by 25-50% over an atmospheric column of ~100km - so over 50m we're looking at "losses" that are right at the limit of the Pancam\NavCam's sensitivity (1 part in a couple of thousand which needs 11bit resolution or better). We certainly see distance effects from Tau and may well be able to see increasing dust levets affecting the view of the far rim during the summer but that dust haze effect is clearly not the same thing.
The wavelength\diffraction limit cannot affect cameras like these in this way - we're looking at macroscale effects not microscopic ones.

The only explanation that makes sense to me is that all of the relevant images are using some form of lossy compression. Possibly someone can find L2\L3 etc +LOCO on the planning data or the analyst notebook that puts the lie to that but I haven't seen anyone point out a set of images with an L2 LOCO and an L7 LOCO of the same spot to show that there is clearly something else going on.

[um3k]

Don't forget that the Pancams use lenses (as opposed to mirrors), so they most likely suffer from some degree of chromatic aberration, especially considering the wide range of wavelengths they are used for.

[fredk]

Here's my stab at putting this optics discussion "into focus".

Helvick is right that scattering from the atmosphere is completely negligible over short distances, where we clearly see that L7 is much sharper than L2 etc.

The IR L2 etc frames are less sharp because the optics are diffraction limited at those long wavelengths. As Bill wrote, the resolution is wavelength dependent in this case, and in fact gets worse with increasing wavelength (essentially the unlocalizable wave character of light is more noticable as the wavelength increases, and the particle/ray character as it decreases).

The optics may not be "perfect", or diffraction limited, at short wavelengths (L6,7), so L7 may not be sharper than L6, but they're pretty darned close, and the point here is that they're both much sharper than L2 etc.

Chromatic aberation may be a factor too, but pancam is pretty slow, f/20, so it's probably not dominant.

As far as "graininess", I haven't noticed any consistent difference between filters.

CosmicRocker, who chewed your backside off? Are you OK now? Can you recall where that was?

[helvick]

Hmmh - it seems my gut feeling about whether the Pancam's could be diffraction limited was just a wee bit incorrect.
Pancam PDS label document

So it appears to me that the combination of f/stop (20 !), wavelength and CCD sensor pixel size (12 microns) means that the Pancam's operate at their diffraction limits.

.. At 430 nm, the actual MTF values cluster just below the diffraction-limited value of 55% at the Nyquist cutoff frequency. Similarly, for 750 nm, the actual curves cluster just below the diffraction limit of 25% at Nyquist. And for 980 nm, the curves drop to just below the diffraction limit of 8% at Nyquist.

I don't know what each of those means but using the same data to calculate the Airy spot size I find that the L2/L3 filter wavelengths diffraction limited airy spot size covers a block of close to 4x4 pixels while at L6 it's <2x2 which kind of ties in with the above quote.

[diane]

Aside from difraction, lens focal point can vary by wavelength. Back in the days of manual focusing, good camera lenses had a different focal mark on the lens focusing ring for visible light and for infrared (I used to do some infrared photography many years ago). It's also the reason why, if you look at something bright red on a bright blue background, the red seems to "pop". That's your eyes having to refocus between red and blue, even if the image is flat like paper or a computer screen.

[fredk]

Thanks for the link, Helvick. Yeah, I forgot to say that it's the small aperture that is really the source of the diffraction-limited behaviour. It's harder to focus a longer-wavelength wave through a smaller aperture.

[CosmicRocker]

Well, it really wasn't a painful chewing. I was just kidding about that. But iirc, I was promptly corrected by a few folks. I'll try to find that discussion, because I recall that the concensus of the experts at the time seemed to be somewhat different from what is now being said.

Thanks to the Google site search function finding the converstion was easy. If I am understanding what Emily quoted from Jim Bell, the L1 (unfiltered) images have the highest resolution.

[fredk]

Diane, that effect is the chromatic aberation others have mentioned. For pancams it is negligible because they are stopped down so much, to f/20.

CR, I've seen that mentioned as well, that L1 is sharpest, but I'm sceptical, since L1 response peaks near L2, where diffraction is clearly an issue.

[diane]

I've always understood "chromatic aberation" to refer to the fact that color focus is not consistent across the entire focal plane. What is sharp in the center is not sharp towards the edges, a that loss of sharpness varies by color. A lot of this is also due to the fact that it's hard to design a lens in which the focal plane is truly flat for all colors.

What I referred to is that the focal plane for different colors isn't even a single plane.

If I'm wrong, please tell me where....