fool moon (typo intentional ;)

This was first attempt at capturing moon, with this frame being thesharpest of a series of a dozen shots or so, but it still is far fromcrisp. Shot with the Canon 100-400 zoom @ 400mm, 1/20s, f/5.6 (sowide-open; I tried different apertures too, but they came out worse),ISO 100. The rig was laying on a pillow (though would dump vibrationsbetter than when tripod-mounted), self-timer, and I think I evenemployed electronic first shutter curtain trick, a.k.a. 'silentshooting' mode in Canon's lingo, instead of mirror lock-up.

What could I have done differently to achieve better sharpness?

Please have a look at Michael Covington's Electronic first shutter curtain on Canon DSLRs: What's better than mirror lock? -- that's the technique I've mentioned in my critique request, and the sharpness I was after.

This is not a bad shot at all.  I find a bit of unsharp masking will go a long way with this king of pictures.  There are a number of softwares like Iris and Registax that allow stacking multiple frames of the same subjet to remove effects of air turbulence.  It will also greatly reduce image noise so you can pump up the iso a bit to get a shorter exposure time.   Iris has a "vancittert" function that does nothing short of a miracle on lunar or planetary pictures.  Unfortunately I haven't been able to make it work with Windows vista.  Look at this thread (in french but it's the pictures that count!):

http://www.parlonsphoto.com/viewtopic.php?p=950036#950036

Regards

François Quesnel

I've had a look at RegiStax when someone mentioned it in the Suppressing noise by stacking multiple exposures thread, but I'm on a Mac and I quickly discovered that it's a Windows-only app (found a good tutorial on it though ;) and unfortunately IRIS is no different with this regard :(

My intention was to use PhotoAcute in the super resolution mode, but I must have done something wrong since only one of a dozen shots came out sharp-ish, so no image stacking for me.  I definitely should have jacked up the ISO to close down the aperture by a couple of f-stops, and perhaps cut the exposure time too.

What you see has already been sharpened; any more and unsightly artifacts become prominent.  However, the large version is a 140% magnification crop, i.e., I've upresed in ACR during RAW development (the native resolution is 5184 x 3456 = ~18 Mega pixels; ACR can produce files with pixel dimensions of 6144 x 4096 = ~25 Mega pixels, hence ~40% increase) as I wanted to obtain as large an image as possible and had only 400mm focal length lens to shoot with.  Downsizing it improves the perception of sharpness, but it's still a far cry from being razor-sharp.

http://static.photo.net/attachments/bboard/00X/00Xmtq-307951784.jpg

 

P.S.
If you're into Matlab -- too technical for me -- there are some Super-Resolution GUI implementations for it (might be worth checking out).

 

Thanks for the info, I'll have fun looking into it.  Reading the exif on your picture I noticed that at the day and time it was taken the moon must have been very low on the horizon.  The reddish tint of the moon and the wavelets on its edge also indicate strong turbulence and high refraction.  Waiting for the moon to get higher up in the sky (above 30 degres) will help in a couple of ways: it will bring turbulence down and make the moon brighter.  1/20 s at f5.6 is a very long exposure for a full moon.  Click on the link below to see a 1/200 s f/12 at iso 200 picture. 

http://photo.net/photodb/photo?photo_id=12040214

Good luck

 

Wow, 1/200s @ f/12 and ISO 200 -- how's that possible?!  Guess you could have opened the aperture up by a stop or two without sacrificing sharpness, jack the ISO up a notch, and you'd have been able to hand-hold the shot.

• Oooh, that's where that ugly CA-like fringing came from...
  ...it was much worse, but I've managed to remove most of it in PS
• I'm such an ignorant it's amusing! :D
  Didn't have a clue about turbulence and refraction affecting celestial bodies positioned low above the horizon; just like the moon most as it rises and has this deep orange colour.
• Yeah, I was able to aim at it while having the rig resting on a pillow on a window sill, so it couldn't have been high up the sky.
• Be careful with the EXIF time stamp: you got it right only by an accident -- I have it set for New Zealand (where I currently live), but was in Europe (visiting) and didn't remember/think about changing the date + clock, but since the time difference was 11~12hrs, 09:28 [am] still makes sense ;-)

I was looking into deconvolution as yet another avenue to increased sharpness.  The most popular (non-astro specific) is Focus Magic, but I'm out of luck with it for now as an Intel-Mac version will be in the making in some unspecified future (funny enough, Acclaim Software, who developed FM, is a New Zealand-based firm :)

Then I read about the adaptive Richardson-Lucy algorithm; apparently Mike L. Unsold's ImagesPlus has that in the toolbox (mind you, someone was claiming its far superior to RegiStax!), but I'm out of luck again: Win-only. Stark Labs' Nebulosity might be my only option as, for a change, it does exist in a Mac version too, but don't know yet what it can and cannot do; it's supposed to be less feature-rich as compared to ImagesPlus, but easier to learn/operate, which for me is a BIG plus (pun intended :)

Finally, though you might be interested in Joseph M. Zawodny's Imaging & Image Processing 'technical essay' on Flickr; I hardly understand a word of what he wrote, but figured you'd be able to make more sense out of it.

Enjoy!

 

Wow! thanks for all that info!  That deconvolution algorithm seems perfect.  I'll have to try it.

Good luck on your side

 

 

Here's one of tonight's blood moon (unimpressive resolution / sharpness, but cool colours ;)