So I'm learning about taking pictures... mostly by screwing up.  I'm learning that brilliant midday sun is generally not great to shoot by, at least for the stuff I am trying to shoot.  I'm learning I really really need to stop saying "ah, I'll just leave the tripod in the car, I won't need that".

I'm learning about masking in PhotoShop, and how to use it to airbrush a subject (thanks James for the link), or how to use it with layer blending to rescue a picture that otherwise comes out half overexposed and half underexposed, like the one at right.

Another thing I am learning about is photo composition, what makes a good photo, and what doesn't.  The most basic rule that I keep hearing about is the "rule of thirds".  This rule basically states that you should divide your viewfinder up into 9 equal sized rectangles (basically like a tic-tac-toe diagram), and try to place your interesting subject matter on one or more of the intersection points as opposed to dead-center in the photo.  Theoretically this makes for a more pleasing photo.  Judge for yourself... which is better?

According to the photo pros, the one on the right is far superior.  The top-left daisy falls on an intersection point, and the out-of-focus small daisy falls close to one.  Of course I am also taking advantage of another "rule" in the photo at the right, "groups of three are most pleasing".  Unfortunately I don't always frame my shots this way because I am still learning.

But thanks to photoshop, I can deal with that by cropping the photo.  I built a template that looks like a big tic-tac-toe grid where the squares are transparent.  I simply paste this over my photo in pshop and then resize and reorient it until I have the interesting subject matter on an intersection point (or on a "thirds line"), and then I crop the photo to that rectangle.  Because I start with a grid the same size as my photo and I shrink it proportionally, the resulting shot retains the same image proportions so the cropping is not so obvious.  Then I just delete the layer which has the rectangular grid in it and my composition is repaired.

Another thing I've been learning about is "depth of field".  When the camera shutter opens you have control over how wide it opens.  This is called the aperture.  The wider the opening, the more light gets in (and of course, the faster your shutter has to close in order to avoid overexposing the photo.)  Another interesting effect of the aperture is how much of your photo is in focus.  A very wide aperture leads to only a very narrow range of distances from the camera where things are in focus and a very narrow aperture leads to an extremely large range of distances where things are in focus.  If you are taking a shot of your subject and there is a mountain in the background, you use a wide aperture to keep just your subject in focus, and a narrow aperture to focus both your subject and the mountain.  The range of what is in focus is your "depth of field", and the different aperture settings are referred to as "f-stops" (as in f/1.4, f/2.8, etc...), with larger numbers referring to a larger depth of field (and inversely, a smaller aperture).

For portraits and macros (extreme close up shots, like the daisies above), a wider aperture is recommended  (I used f/4.5 on the three daisies pic, which is why one of the daisies is out of focus).  Using a wide aperture will blur the background so that your subject stands out and becomes the focal point of the picture.  A nicely blurred, even, and nondistracting background seems to be the best, unless it is intended to convey context.  The flickr folks toss around the term "bokeh" (Japanese for "blur") in reference to this effect.

Speaking of backgrounds, you have to think about them a lot.  If your background has too much going on in it, or has a major distracting element, it detracts from the photo.  When I first took the picture of this rugosa rose covered with water droplets I was sure I had a great shot.  But the barn in the background was distracting and various commenters on flickr told me so.  Unfortunately this shot can't be saved by a crop if I want to see the whole rose, but it is a reminder of a good lesson for me, so I hang onto it.

The one exception (as noted earlier) is when the background provides useful context.  If your subject is a guy sitting in an outdoor cafe with a big boat oar leaning against his table, the out of focus image of Venice in the background communicates useful information!

I seem to be pretty good with macro shots.  And some of my landscapes come out nice (and more would if I would start paying attention to the frikkin lighting).  But so far, I'm not good at portraits.  This is mostly due to lack of practice... nobody I know really wants to be photographed hundreds of times while I figure out how to do it right, it's awkward to photograph strangers on the street (but apparently legal if you don't profit by it), and I can't afford to hire a model.  But sooner or later I'll figure something out.  I've taken precisely 2 portraits to date where I was pleased with the result, and only one of them is of any quality.  I definitely need to work on this department.