The Integral

Photography is used by humans for various reasons, but most of all, to capture special moments and to tell stories later on. How well the photograph tells the story, reflects on how well the photograph has been taken (or edited). Point and shoot cameras are perfect for that quick shot with 4 sets of shiny teeth in a popular restaurant, or that cute family picture on the beach. Often, what you see is not what you get! And this is where a DSLR comes handy, particularly for a few features such as ability to shoot in RAW, ability to manually adjust exposure and thirdly (but not lastly) the ability to bracket images of different exposures. Oh, there are a few excellent P&S cameras out there that do all of them, but these functions primarily belong to DSLR cameras.

Which pic do you prefer: the left one, or the right one?

Have you ever taken a pic of inside a dark room with bright sunlight visible outside a small window? Have you ever taken a pic where there are beautiful clouds in a bright sky and your friend standing under the shade of a canopy? You would often observe that either the sky is blown off (too bright to see any details in the clouds) or the subject is too dark. If you have a DSLR, you would often find yourself asking the question “should I meter the bright sky or should I meter the dark subject in the shade, and still preserve all details”.  Interesting thing here, is that you would be able to see the details in the sky and your subject perfectly fine, but your camera just wont be able to capture them. The problem here, is not with your photography skill, but a limitation in the camera. Let me explain:

The illumination ratio between the darkest spot and brightest spot (containing detail) is called the Dynamic Range of a scene. Typically, DSLR cameras can capture contrast ratios of around 2000:1 or 11 EV (when taken to a base of 2), LCD displays can display upto 9.5 EV. The Human Eye, can perceive well beyond that range. However, there is a catch. The human eye cannot perform both operations (viewing very bright and very dark details) simultaneously, this is done by adjusting the amount of light that enters the eye and it takes a short while to adjust. A camera cannot do this, as you have just one moment to capture, just one frame of details. If you open the shutter for too long, you get too much light (and the bright areas get blown out), if you open it for too short, you get too little light (and the dark areas remain dark).

To overcome this, photographers use multiple exposure-bracketed images to create one single image with High Dynamic Range. Are you getting the hang of it now? In principle, take three images - 
Pic 1: A little darker, so that details in the bright areas are maintained (-1 EV). The darker areas in this image would be blank with no details.
Pic 2: A little brighter, so that the details in the dark regions (shadows) are clear (+1 EV). The bright areas in the image would be blown off.
Pic 3: The last pic with an average exposure of the two above  (0 EV). Taking just this pic would not help in situations of extreme contracts, such as bright exteriors and dark interiors.

I took this set of bracketed pics at the Baha’i Temple in Wilmette, Illinois (an affluent village 14 miles north of Downtown Chicago). This was the first of the seven Baha’i Houses of Worship currently surviving. The first pic is –1 EV, the second pic is +1 EV, the third is 0 EV. (Click on them to see large)

Now put the three pics on top of each other in any image editing software and use the bright portions (highlights) from the first pic, dark portions (shadows) from the second pic, and the mid-tones from the third pic! Thats about it!

It sounds straight forward, but to do that, you need a few things:
1. Take exposure bracketed images of exactly the same frame – use a tripod and the AEB feature of your camera. Turn off auto-focus!
2. Use a photo-editing software to put the three HDRs together. I am not sure if you can do it in Photoshop, but there are plugins you can install.

Alright… If you have come this far, it makes some sense to go a bit further…

Perhaps, you did not realise, the method I walked you through to create an HDR, will still not make your images look stunning. Reason: Your LCD monitor can display images only upto a certain contrast ratio. Its dynamic range is lesser than your camera’s standard dynamic range. Now you have an image with dynamic range much greater than that of your LCD monitor and even your camera. So what do you do?

You tone map the image! Basically, you reduce the contrast ratio of the image from something like 16 EV to about 8 EV, or, from 100000:1 to about 255:1. The JPEG images you see usually, are 8 bit images, that contain upto 255 levels of illumination (from darkest to the brightest). So the goal of tone mapping is to reduce the contrast ratio but to preserve the detail. This, can be done using two techniques – Global Tone Mapping, or Local Tone Mapping.

Global Tone Mapping is the most basic operation. It is similar to contrast reduction in Photoshop, or even your LCD monitor! It reduces the contrast in each and every pixel of the image, uniformly according to the most appropriate function for that image. This, does not help us in what we wish to achieve. Really.

Local Tone Mapping reduces the contrast of each pixel based on the contrast and illumination of the surrounding pixels. If you have studied Calculus/Derivatives in school, you could think of it to be something like Global Maxima / Local Maxima. Local Tone Mapping on an image having shadows and highlights would have different impact on the dark areas and the bright areas. In fact, it would bring out more details than you thought were present in the scene! HDR tone mapping using Local Tone Mapping has become extremely popular amongst photo enthusiasts for exactly this reason. These photographs have a distinct look and seem attractive to many! Often one would get comments like “This looks so much like a painting”, or “This is like a computer game scene”. Limited dynamic range is never a problem in paintings, whereas computer game manufacturers use tone mapped HDR graphics to improve the look and feel of the play. While HDR in video games make scenarios look real, HDR and local tone mapping in photography makes real scenes look a bit graphic!

 
HDR created by merging the 3 bracketed images above. Minor tone mapping applied.

Major tone mapping applied to reveal extra details. Looks more graphic than the one above.

Sorry about the long blog post :) I should keep myself in check so that I do not go overboard and do too much of local tone mapping. My name itself says adi… which means, excess!

Recommended softwares:
1. Adobe Photoshop / Lightroom
2. Oloneo Photoengine (the one I used)