10 Questions on HDR

Tech Info: Nikon D300s with Tokina 12-24 lens @12mm, ISO 200, f/5.6 for 1/1500 sec

A day long skiing event, coupled with long flights has taken a toll on my disc. I’m pretty much tied to the bed lying flat to give some rest to it. Sad news is that I won’t be able to play the weekend game. Also I’m not in any mood to process my pictures. So I thought will use one of my old pics to answer some questions which are asked by my friends. Hopefully it’ll be useful to you and you can get the jargons cleared when people talk about HDR pictures.

What is HDR? Its an acronym for High Dynamic Range 🙂

Why HDR? Modern cameras are not equipped to capture all the information of a scene which has a high contrast in hightlights (bright areas) and shadows (dark areas) of a picture. Most of the time, you’ll have to risk having a blow-out areas (completely white or black) in your pictures. So HDR imaging is a way to capture all the details of the scene by using multiple images and then blending them together.

What is tone mapping? We don’t have any monitor (that a common man can acquire)  which can display an HDR image. In order to display those images, we got to compress that large dynamic range into an 8 bit image which can used to displaying on the web. Tone mapping is the name given to that method.

How do you make an HDR image? Like with any new imaging, there are multiple ways to get this done, but the most common method is to use the built-in feature of modern DSLRs called Automatic Exposure Bracketing (AEB) for your shots. The camera automatically takes shots with +/- EV values based on your setting which can then be used to generate a HDR image.

What’ll be the workflow for HDR? After you have your bracketed set of shots, you feed them to a tone mapping software program like Photomatix to generate a .hdr file. Then you would further compress that hdr image by tone mapping using the available options with the program. The resulting file would be a tone mapped image which can either be used directly or further processed depending on your taste.

What do you mean by those numbers? When you bracket your shots, you’ll have a base shot which is seen as the correct exposure by your camera. Now this wouldn’t be the capture of the complete range, so you would start capturing more frames by under/over exposing from the base value. This is usually done by varying the shutter speed. So if the 0 EV image (the base image) is at 1/1500 of a sec, then a +1 EV would be one stop more ie 1/750 of a second. You would continue this to get a number of images. Usually people take equal number of frames on either side to get the complete range. So in the above picture, I got 5 frames starting from +2, +1, 0, -1 -2. The exact number of images would largely depend on the picture.

How do you decide on the number of frames? You can take the complex route of determining the luminosity values of the areas in your picture, or simply use the camera’s histogram to aid you in determining the number of shots. Typically you want the histogram to cover the three areas (shadows, midtones and hightlights) with good amount of detail without clipping on either side. So take as many shots to cover that entire range.

What tools do you need in the field? You would definitely need a camera with custom controls. A simple point and shoot camera with pre-definited control knobs only won’t just do it. The camera needs to have either Aperture Priority or Manual mode. Of course a tripod would definitely help a long way. You can try doing it hand-held by using the burst mode in the camera. But a tripod goes a long way in making sure the elements of the scene are aligned properly. A scene with more static objects are suitable and that also helps in getting the pictures aligned  properly.

What programs can you use for HDR? Any software that can aid you in combining images can be used to a dynamic range in the picture. But the most widely used software is Photomatix. You can also use the open source program Qtpfsgui to get that done.

How did you process the above image? The tech info above mentions the 0 EV frame. The five frames used was from +2, +1, 0, -1 -2. After generating the tone mapped image, I used it to layer over the 0 EV frame and did some work on revealing the underlying 0 EV using a paint brush.

The following are some of my HDR images. I’ve shot quite a lot of them, but haven’t yet got to processing the remaining ones. Some day will get them out 🙂