HDR and the Pursuit of Mystery

Many photographers lately have experimented with High Dynamic Range (HDR) photography.  This technique involves taking multiple exposures of the same scene but with different exposures to capture the full range of light in the picture.  These are then blended in software and possibly tonemapped to produce an image with a wider range of brightness than any one image.

Your eye has a dynamic range of about 16 million:1.  In contrast, a digital camera has a range of about 256:1.  What that means is that you can see a much wider range of light in a scene that what can be reproduced in a camera.  To overcome this limitation, we can use today's digital cameras to stretch their range.  This involves taking multiple exposures.  Typically this involves three, a normal exposure (0ev), a dark exposure (-2ev) and a bright exposure (+2ev).  The theory is that these three pictures should be able to cover the full range of light in a high contrast scene.  These three exposures can then be blended in post processing to produce a picture that captures a wider range of lighting.  In essence, you can now see the darkest and brightest parts of the picture.  These are the blown out white skies and/or totally black foregrounds that show up in so many pictures.

These techniques can be used to produce surrealistic effects.  One website that has promoted HDR is Stuck In Customs.  Going there you can see how he processes his images and produces a particular style.  HDR processing can produce unrealistic results with overly vibrant colors, halos, moody skies, and lighting seeming to come from multiple directions at once.

For those who haven't completely fallen asleep, I am really going somewhere with this (I think).  Personally I have run through the gamut of HDR processing, starting with the high contrast vivid colors that look like they were taken on planets with multiple suns to my currently more realistic bent.  The first picture is a sunrise in Monterey, CA.  I started with three images, preprocessed in Photomatix and then continued in Photoshop.  Bright colors with the shadows filled in substantially and it looks like there may be another lightsource to the right with the way the foreground is lit.  This is tonemapped in Photomatix with further adjustments in Photoshop.

But I have been experimenting with blending the exposures by hand.  The second picture uses the hand blending.  This doesn't have as strong a color set and the contrast is toned down in the front.  The vertical rocks have substantially more shadow which is what I was trying to accomplish.  I wanted to retain the dark shadows but produce enough detail that the viewer could see some of what was hidden.  Currently I like the second better.  The reason is the sky shows more subtlety in the colors and the shadows are more pronounced.  But I did dodging in Photoshop to bring up the shadows somewhat but not completely.  Just enough to show a little detail and then stop. 

My point (if there actually is one) is that I'm trying to convey the scene but give the viewer a glimpse at all the detail that may not be evident even when standing on the shore.  That requires HDR but I don't want it to look like I've lit the scene with a nuclear powered flash unit.  I wanted to retain the darkness but still show the details.

I have been reading a book lately that actually says in a coherent fashion what I was trying to do with the recent pictures.  It's "Landscape Beyond" by David Ward.  Besides talking about taking 'small' landscapes he also has a chapter titled "A Sense of Mystery".  If you can get it from your local library take a look.  There are some interesting ideas.  But his point (made directly) and my point (made in an extremely roundabout manner) is that a photographer should try and maintain a sense of mystery in their work.  This requires making images more than taking images.  To leave some things unsaid or somewhat hidden so that the viewer must engage in the picture and come to grips with the ambiguity or inability to see everything.  The viewer then adds their own interpretation of the image.

If you have an opinion, let me know.