Tag error: <txp:tru_tags_if_has_tags> ->  Textpattern Warning: tag does not exist while_parsing_page_form: archive, default
/commentable: Making a "good" photograph
Dinarius = digital interest
13 February 2012

Making a "good" photograph

What makes a good photograph? What’s a good picture? That’s a valid, if vague, question. Aesthetics aside, there is a real answer. As a photography teacher I get the question a lot. Posting this may seem simple but it’s a revelation to the newer shooter.

What makes a good picture, technically?

The Histogram illustrates something that black and white photography students may remember: Full Scale and Full Range. These were the hallmarks of what amounted to a “good” picture. When my teacher checked for Full Scale, she was making sure that the photograph had a black and a white. Full Range was more values of gray. We averaged about eight grays; college folks could manipulate closer to twelve different grays and really good photographers could squeeze upwards of twenty-one shades of gray.

The more grays, the more SILVER the print looked; it was beautiful indeed. Now with color and digital and a billion ways to share instantly, this all got lost. Whether black and white, color or either in digital, Full Scale and Full Range are still the ways to go. The advantage with digital software is that it SHOWS you how you’re ranking in these departments.

So, in short, when you’re shooting and you want a picture that will afford the most flexibility for post processing, you can use your in-camera HISTOGRAM to find the shot that gives you the best BELL CURVE. Here, I used GIMP to provide a Histogram of Values. I found that a neighbor has a great tree. So it modeled…

As the camera balanced the scene, the curve is low.
I would say to you, here’s a picture of a cool tree. And this is just how the Canon 50D balances it. This is the Camera Balanced Curve. Dark: Weighted to the left, and; washed out: Weak range on the right side.

Better histogram from Average HDR of three images
Using yesterday’s technique to Tone Map THREE images of that tree, I would have a richer example and better Curve. By “RICHER,” I mean, technically too dark. Aesthetically fine, but amounts to loss of detail.

HDR and Tone Map treatment produces a near bell historgram
Bracketing that tree results in a much better Curve and you would finally “see” the tree nearly as it really is. This isn’t Full Scale – the Black isn’t 100% – but below, you’ll see how much more fidelity this has all around. That’s the Bell Curve in the Histogram for you.

It’s only occurred to me after posting that I’m using High Dynamic Range software which is sort of cheating since I’m supposed to be showing you the camera’s ability to improve your image’s range. Meh, this will happen… I was more interested in knowing I could fetch some Histograms for you without too much fuss.

Side by side by side

The Camera Balanced Curve lacks subtle hues, washes out lighter details and the darks just seem muddy when compared to the other examples. We know that HDR software will bring out more Range (the HDR image actually lacks Full Scale because I didn’t tweak the Levels enough), but your camera’s settings will also improve the results of the Histogram Curve if you re-balance its light while you’re shooting.

In a case like this, you can push the exposure to allow more light to come in. With the wind today, a wider aperture would have been better than a longer shutter speed. Once the bell curve edged toward the right, the bright side, we would have an image that is considered technically “good” and quite flexible to work with in any software.

Closer look at the histogram bell curve

Favorite's the ARTICLE, not the SITE.