High Dynamic Range or better known as HDR, has been a concept that seems to be so very misunderstood. Hundreds of web pages with 2D and 3D designers trying to understand and justify this paradigm shift that will change the way they think about photography and lighting in general.
HDR is not a concept that is just a fad or some crazy idea that only the highest end professionals ever have to think about, but rather something that will change everything from how you take pictures with your family camera to the way that the Hubble Telescope takes pictures of our sun.
If we take a simple observation of how light looks in the world around us, you will notice that there are areas that are too bright for you to look at like the sun, and areas too dark for you to see anything. This idea refers to the RANGE of light (or sometimes referred to as luminance) that is visible to the human eye. The human eye can see a range of approx. 10,000:1 but this range can adapt over a 1,000,000:1 range (walk outside, your eyes slowly adjust so you can see better at those luminance levels), so here lays the problem. Presently our monitors, cameras and TV’s have very limited range, approx. 100:1, which is 1/100 of what we can see. As a result of this limitation, file formats were created to work with these limited mediums such as Windows Bitmap. Original JPEG and TIFF images could only encode a limited brightness range of 0-255 (i.e. the brightest point can only be 255 times as bright as the darkest point). The move has finally been made after 30 years of these limitations, both JPEG and TIFF file formats now support higher and broader color depths.
The first thing everyone thinks, is why would I need anything more if you can’t see it on a screen or capture it with your camera. This is wrong, and where it gets very technical. In the world of computers a BYTE is limited to values between 0 – 255, so let’s say you store all your image information as BYTES and you brighten your image by 20 value points, any value that was greater than 235 now becomes 255 and all the detail that was stored between 235 – 255 is now lost. So if the same image was darkened by 20 points the image would make any value that was 255 equal to 235 (rather than a range of values between 235 and 255). This always leaves your image with a contrast reduction effect.
To work around this, the HDR (High Dynamic Range) process was developed. High dynamic range image files are special formats that accurately encode the extreme luminance of very dark and very light areas of the scene without any visible data loss. HDR images address two issues, the first is that it can now store values ranging from 1x10exp(+/-37) and the second is that the range is so large that programmers never have to clamp the values of their image to a fixed max value such as 255. So by using the example above if the value was increased and then decreased by 20 points the original values would be restored leaving the image unchanged. This sort of editing is referred to as “Non-Destructive”, since any change can be reversed without data loss.
HDR images are normally created using digital cameras that can record different exposure levels (some cameras even come with Auto-Bracketing), so that they can be combined later to make a very detailed images storing data for both very dark and bright areas. This leaves the user with all the control to decide on how the final result will look.