Quote Originally Posted by OldClicker
I lot of objection from many of those opposing some type of post processing (not just HDR) is because, “It doesn’t look real.” But I find that they have defined ‘real’ as ‘what film looked like.’
TF, in a very real way, that is an accurate comparison. The human eye and film "see" light the same way. A digital sensor does not. If we were to put a 25 watt light bulb in a large room, the eye would see a gradual progrssion from the lightest part of the scene into the darkest. Film would capture that scene exactly the same way, very closely approximating what the eye saw. But a digital sensor 'sees' (and captures) it as a series of half-lives: 50% of the total available information is contained in the brightest stop. 25 % in the second brightest stop, and so on. So in a typical digital image with a 6 stop dynamic range, while you and I see something like this:



The digital sensor records something much closer to this:



As you can see, in the digital image, there is more detail in the brightest stop, but every other stop has less information. This is what makes HDR work well for digital images- by varying the exposure and creating an separate images for each individual stop, you can place that first 50% in EACH range of the image, giving the capability of much more information than is possible even in film.

Now, if one was to do that, then systematically remove information from each successive brightness range, it would be possible to take "reconstruct" the gradual transition from light into dark that film captures.

So, it stands to reason then that it IS possible to take an HDR image and use it to create something that is more "realistic". That is, you can create an image that is closer to what the human eye perceives. The problem is, most of us don't. We don't remove that information, and the result is that over-processed look. Way too much information to be realistic.

- Joe U.