I issue my own accolades for a superb shot. But I am curious about the
choice of f/5.6 on the 24/2.8. With the emphasis on short exposure times
opening up another stop would have made a significant difference. The only
things that need to be rendered sharp are the stars. Was it your judgement
knowing that lens that f/2.8 or f/4 would have introduced visible
aberrations into the image?
Thank you.
I had three issues with using f4 for this shot.
1. Inability to achieve critical focus in a timely manner. I had only
a few minutes of activity and had to act fast. My night vision was
compromised due to a migraine and I couldn't focus, so I did the
normal (from previous experiment) and that was to hit the infinity
stop and back off a millimeter. My first shot, which was wide open and
6 seconds, was a little off, and I used that as a guide to adjust the
focus a touch. At f5.6, the stars are pinpoint sharp.
2. Coma. The coma is pronounced wide-open, and is still disturbing to
me at f4. Normally this wouldn't be an issue, but the sky was so clear
and as it was moonless, the stars were "in your face" bright. I wanted
a good star-scape. You will notice that most aurora photos don't have
much in the way of a star field. It's been compromised out. A better
job at focusing would have reduced coma a bit too.
3. Lens shading. my f2.8 shot was the exposure I wanted (6 seconds),
and would have had the sharpness I needed if I would have been able to
achieve critical focus, but the vignetting was pretty severe. This is
much more an issue with the Canon sensor than with film. I made a lens
profile for the 24/2.8, that addresses the lens shading but at the
expense of raising noise. F4 is better, but f5.6 is almost perfect.
The lens profile also addresses coma a little bit.
I contemplated using a higher ISO. In hindsight, I probably should
have. I'm still learning this camera and what I can get away with. ISO
12800 would have helped with the structures of the auroras better, but
at the loss of color. The higher you go, the less flexibility you have
with color. I knew I needed to bend the pixels a lot, so I needed all
the color information I could get. The 6D's chroma noise is baffling.
I didn't go to 12800 because the image playback looked acceptable. It
wasn't until I got the pictures in the computer that I noted the
extreme loss of structure in the longer shutter speeds. For future
reference, I need to stay under 10 seconds,
I was going to change the lens to the 50/1.4, and try some tighter
shots at f2, but they faded before I had a chance. My window was about
10 minutes total and of that, only five minutes yielded anything
usable. The later shots were technically better (fixed the focus
problems, etc), but the patterns died away. The fifth shot in the
series (this one was #3), looks to be about perfect, but the dancing
pattern isn't as pretty.
AG