holding it together (before and after)

Canon S95 at 28mm, 1/15sec at f2, ISO250

This is the Clark St. Red Line platform.  I titled this photo "holding it together" because I felt the guy in the lower right kind of holds the composition together.  (it also looks like he's holding it together, too).  Without him, this would be a nice, but rather ho hum shot of a subway platform. 

This photo was taken with a Canon S95 and took an interesting journey from start to finish.  Here's the original RAW image, opened in Photoshop and converted to jpeg:



Um, where's the dude's head?  The answer is, it's not there.  Here's what happened: The S95's sensor captures 3648x2736 pixels (which is a 4:3 aspect ratio), however, it gives you the option of capturing at 16:9, 3:2, 1:1 or 4:5 instead.  A cool feature, and I've been using 3:2 (3648x2432) because that's the same ratio that my DSLR records.  I assumed the S95 discarded those extra pixels, but it turns out I was wrong.  Here's what happens when I open the RAW file with DXO Optics Pro (instead of Photoshop):



Bingo, the full 3648x2736 is there!  Cool.  I'm not sure why this happens, but clearly whatever Canon is doing to delete those extra pixels is being ignored by DXO.

Now, normally this would not be very helpful.  If I'm composing for 3:2, I'm composing for 3:2 and I don't want the extra pixels on top or bottom.  But in this case it worked out-- I knew immediately I had mis-composed this photo, but the guy in the right corner moved before I could take another shot.  The extra pixels in this case helped me create an interesting image out of what would have otherwise been a throwaway shot.