Rev 123 - Dec 18, 2009-Jan 3, 2010 - Prometheus Options

[Hungry4info]

The image is terribly noisy, but you can see the shadow of the moon in the F-ring.

http://saturn.jpl.nasa.gov/multimedia/imag...6/N00148979.jpg

[S_Walker]

[quote name='ugordan' date='Dec 27 2009, 01:08 PM' post='152478']
Awesome Prometheus images, a quick RGB composite from the first set:

Very nice. I was wondering what you're basing your color balance on. I did an RGB combine after cleanup and alignment in MaxIm DL5, and I end up with slightly more color variance across the little icy rock. My version also uses a touch of non-linear stretching so that the bright regions are easier to see. Upsampled and slightly deconvolved (0.5 PSF, 4 iterations).

Sean W.

[ugordan]

I was wondering what you're basing your color balance on.

I based it on the F ring white point which in principle should have also resulted in Prometheus' color being close to some calibrated shots I have (this one and this one) - a slight brown-reddish color. Mind you, I did this on a laptop screen and I have no idea if it turned out remotely close to what I wanted it to. There definitely is some slight color variation, but just like in Hyperion's case, if you don't "kill" the overall hue, it's not noticeable. Otherwise you just have red and slightly-less-red regions.

Upsampled and slightly deconvolved (0.5 PSF, 4 iterations).

Do you use any special tool for that? I'm asking because I'm looking for a more specialized deconvolution tool than the "smart sharpen" filter in Photoshop for Deep Impact imagery.

[S_Walker]

Do you use any special tool for that? I'm asking because I'm looking for a more specialized deconvolution tool than the "smart sharpen" filter in Photoshop for Deep Impact imagery.

Thanks, I'll try that. I use MaxIm DL CCD 5.08 for resampling and deconvolution. It includes Richardson-Lucy and Maximum Entropy algorythms, though R-L (or as others call it, LR) is the better version, based on the code developed to fix Hubble images before COSTAR.
http://www.cyanogen.com/maxim_main.php
You can download and try the program free for 30 days. It is pricey though. Another alternative is a plugin for Photoshop called Focus Magic. Google it. It works well also, though not nearly as powerful.

I use LR often with my amateur astrophotography to deal with my usual poor seeing conditions:

http://masil-astro-imaging.netfirms.com/Solar%20System.html

http://masil-astro-imaging.netfirms.com/Latest_Images.html

[elakdawalla]

I cheat to get color. RGB color combos from Cassini images almost always have some completely white pixels due to the automatic stretching. To set the color I just go grab something Gordan made from calibrated imagery and set the RGB values of that brightest pixel to those of a bright pixel in a Gordan image. It's not science but then images made from raws aren't science either.

Also for the version I posted in the blog I made an RGB combo from red, green, and blue filter images, and also stacked the red, IR, clear, and green channel images to make one that was less affected by JPEG artifacts and noise, then did a nonlinear stretch on it to bring out some more of the detail in the brighter regions, then sharpened it just a little bit. Then I converted the RGB to HSB and swapped in the sharpened image for the clear channel.

[S_Walker]

Thanks for the insights. This is probably closer to accurate then.

[ugordan]

Which is kind of similar to what I did above - selected a small rectangle (for better s/n) in the thick of F ring, removed "dust and scratches" and averaged it and thus got the white point and channel balance, seeing how much each channel needs to be boosted to make that averaged portion gray. In 16 bit mode, of course, every little bit helps. Then just applying those factors to the entire composite. Obviously only works if you have something you know is pretty much white/gray in the source image AND is not overexposed in any channel - F ring, Mimas, Enceladus and leading hemispheres of Tethys and Dione work well here.

When it comes to raw jpegs, we all cheat. Only our methods may differ

[S_Walker]

Which is kind of similar to what I did above - selected a small rectangle (for better s/n) in the thick of F ring, removed "dust and scratches" and averaged it and thus got the white point and channel balance, seeing how much each channel needs to be boosted to make that averaged portion gray. In 16 bit mode, of course, every little bit helps. Then just applying those factors to the entire composite. Obviously only works if you have something you know is pretty much white/gray in the source image AND is not overexposed in any channel - F ring, Mimas, Enceladus and leading hemispheres of Tethys and Dione work well here.

When it comes to raw jpegs, we all cheat. Only our methods may differ

Though we tend to come to similar conclusions. If you decide to try MaxIm, feel free to ask me any questions you'd like via my Sky & Telescope email address.

Happy new year!

[jasedm]

Very interesting first images from a reasonable range. Prometheus is much closer to Pandora than Atlas in a superficial visual sense.
I was expecting to see much more ring-deposited material - at least on the anti-saturn end of the moon, especially since the satellite regularly dips into the core of the F-ring.
Nice to still be surprised by the Cassini results.

Jase

[JohnVV]

Why would anyone foul a picture like that up with Star Trek stuff?

Alright everyone, please, enough of the Star Trek image discussion. Jeez, I wish I'd never said anything. That'll teach me crack a joke.

as long as things like this are at a min. and do not take over a thread ...
there as always a bit of room for some late night ( 4 am) fun . but i did like the Santa and raindear in front of saturn .

[Astro0]

One of UMSF's alumni makes a contribution in this MSNBC article

[nprev]

Right on...congrats, Emily!!!

[tedstryk]

Awesome!

[Phil Stooke]

I hate being on vacation - I can't play with all these new goodies!

Phil

[Ian R]

A crescent Saturn, taken on the 2nd of January:



Happy New Year to everybody!