If you have been into photography long enough to have developed film, you will remember that one of the tricks of the trade is to subspose the film because, in the dark room, it was always easier to dogde (lighten) and retaining detail than burning (darken).
I have not been around that long, and although I did have a go at B&W processing, to be honest, I was more focus on achieving a positive image at all, forget fancies such as dogding and burning. However, one of my first books on photography, when already into digital, was a 1950s manual I got at an old bookshop. The basics could not be that different right? Well, I was wrong.
The manner in which a modern digital sensor captures light could no be more different that the way that photosentitive film did (does).
Permission to get a bit tecnical (just a bit) and I note of caution, this is only applicable if you are shooting RAW.
When shooting RAW (always), my good old Canon 40D, captures 14bits per channel (there are three channels, red, green and blue, irrelevant for this explanation). That means there is a potential for up to 2^14 tonal levels (16,384 exactly) and (I read) a Dinamic Range of 9-stops. One could assume each stop range would be able to enjoy its proportional share oft the tonal levels (about 1,820 each) but, unlike the more equalitarian film, that is not the way digital sensors work.
The brightest stops ‘gets’ half of the available tonal ranges (8,192), the second stop-range, half of the left over (4,096) and the process continues so that the last stop-range, the darkest, can only enjoy 32 tonal ranges. Your dark range can hardly retain any detail, no matter how much dogding, there is hardly anything there.
So.
Everytime you expose aiming to get a well distributed histogram remember that the left side (the dark side) is capturing an exponentially smaller amount of information (tonal range).
When I read an article about this (apologies I cannot atribute correctly, this is years ago) and ETTR, Exposing To The Right and my photographic results were changed forever.
Just as it says on the tin, get used to expose aiming to “fill up” with data (mind, without clipping!) the right side of your histogram, remember that way you are capturing as much information as possible, even, from the darkest areas of your scene.
No pleasure without pain……
1. You will need to get used to read the histogram on your LCD back, if you have clipping warnings activate them if you need to (I personally do not like it)
2. Your pictures will look dreadful in your LCD back (overexposed and little contrast), they will continue to look as bad on your computer screen and it will take some postprocessing to get there but the results will be incredible. Eventually you will be able to preview the good result you will obtain just looking at the back of your camera.
3. Your files will get “fatter”, remember you are capturing more information, but hey memory cost and buffering speeds today are not an issue.
I took this picture in Gerrans Bay, Cornwall, literally from the window of my car (stationary and out of the traffic!). I really wanted to ensure that I could capture as much detail of the clouds as possible, so in order to get as all possible tonal range on the brightest stop ranges I went 400ISO, ƒ5.6 and added +1/3EV for good measure, sort of crazy for a mid summer afternoon shoot at the sea, but it was exactly as much as I could push the ETTR technic. (It was a fast shot at 1/3200″). Then it was about my normal 1, 2, CVS and 4 workflow .
Have fun and remember not to frustrate yourself with the results on the LCD, that will be the biggest challenge you’ll have. You have been warned.
There are fantastic articles out there on the subject…you can start here…
http://en.wikipedia.org/wiki/Exposing_to_the_right
Ramon








