Dynamic Range Optimization

Over the past few years much effort has be put into trying to make digital camera image quality as good as if not better than that of film. Generally this has been concerned with resolution and noise levels, but another particular area where some effort has been spent, is in attempts to widen dynamic range. Fuji have been at the forefront of this producing cameras with different sensor designs to those normally used, and involving pixels of differing sizes.

The dynamic range that a sensor can cover is a vital part in the production of the best possible image quality because it sets the limits of the tonal range that can be captured in any scene. How much detail can be depicted in areas of shade and shadow, and perhaps more importantly with regard to digital cameras, and where the biggest problem usually occurs, how much detail can be retained in areas of brightness.

In scenes where these problems arise, those of high contrast, bright sunny days where areas of both brightness and deep shadow exist, attempting to allow for these, by either increasing exposure to reveal shadow detail, or reducing exposure to preserve detail in bright areas (highlights), means loosing detail at one end or the other. Shadow detail is only produced at the expense of lost highlights, whilst these are only preserved at the expense of lost shadow detail. The narrower the DR range the worse this problem becomes, and conversely the wider the DR the less of a problem it is.

A recent new arrival increasingly being added to camera specifications, most notably DSLR's, is the option to 'extend' or 'increase' the DR range. This is often implemented at the cost of a narrower ISO range, comes with various different names depending on the camera maker, dynamic range optimization, highlight tone priority, shadow adjustment technology, active D-lighting, etc, but which all works in roughly the same way, and is software based. 

In spite of the rather grand sounding names that have arrived to describe it there isn't really a lot to it actually, for it has nothing to do with sensor design, but post processing of the sensor information. It's one reason it is being so widely incorporated into cameras, for it can be used with virtually any camera and sensor design, and as such doesn't cost much either. 

In fact it's just an automated in-camera method of image adjustment that anybody can carry out separately with any camera and a decent image editor, for basically it's really nothing more than under-exposure to preserve highlights coupled with subsequent levels adjustment to lighten shadow details. A method of preserving highlight detail whilst still producing details in areas of shadow. Our pages on  Image Exposure  &  Using Levels  deal with both these issues, and show you how to undertake it and the results obtained, whilst the page on  Histograms  will probably also be of benefit. 

Having the camera decide automatically when scene highlights will be over-exposed and allow for it in this way, applying subsequent alterations to shadow areas to produce a balanced final image which needs no further alteration sounds good. The problem will be in discovering how efficient each particular camera is at actually implementing this. For deciding at what level to carry out these alterations, by how much to under-expose the image, and at what level to raise the tones of the shadows to, is something that can be very subjective. Because a major side effect of under-exposure is that noise levels can rise, and become much more evident in darker tones if these are subsequently lightened. It needs a very deft touch to find the right balance between the two. But if it is proved to work satisfactorily then it will be an advantage to have and to use.

The one area where we are rather puzzled with all of this in-camera DR range widening, is with the reduced ISO range that often accompanies it, for it is not an adjustment that is concerned with the ISO speed the image was taken at, just the tonal range it has. Many cameras that now offer it that have a range of say, ISO 100-3200, reduce this to 200-1600 when it is implemented. We can't actually work out why this should be so, except to think that it is perhaps somehow tied to camera image processing parameters. That for example the camera under-exposes by 1Ev and then lightens the shadows by 1Ev to compensate, and needs base parameters, tonal exposure values, around which to base all the decisions and tonal changes that are carried out, and has to use those it is given for normal image processing.

Anyway, all we can say is that welcome as these extra image processing specifications are, they shouldn't be viewed as something that you can't do without and are worth changing your present camera for, just a useful additional bonus to have if you happen for other reasons to purchase a new camera.