The dynamic range of a photo is the difference between the darkest and the strongest light in the image (sometimes called extended tones). The important thing to keep in mind before you start is that nature is well done, the human eye has a very wide dynamic range (called visible spectrum or optical spectrum, which covers wavelengths from 380 nm to 780 nm from violet to red, before and after ultraviolet infrared): in other words, looking at a landscape or any scene, we are able to discern the infinite shades between light and the shadows ... Difficult to make both these strong contrasts with a camera and a display screen or paper printing.

In the end, this is not the actual photo whose dynamic range is evaluated, but the media or instrument capture and restores: human eye, screen, paper, sensor digital camera, movie ...

Although we will see that in the end, methods to maximize the dynamic range of a photograph does not begin with the advent of digital photography, HDR techniques as perceived and used today have their origin in digital imaging and computer-generated images (including the work of Professor Paul Debevec in ICT Graphic Lab California in 1997). There followed a whole bunch of software, filters, techniques and other IT artifacts to literally create images with optimal exposure from multiple photos.

You read that right: to create a HDR high dynamic light, you need more input photograph of the same scene differently exposed. The principle is actually very simple. Indeed, sometimes our photos are underexposed, sometimes they are a bit overexposed, sometimes very over-exposed, etc.. : Why not take the best of several differently exposed photos to create the "perfect" show? That's the HDR.

Technically, "take the best of several differently exposed pictures" is not so simple and it is not possible natively with our usual file formats 8-bit per channel (256 shades) or 16 bits per channel (65,536 shades) . This is why the HDR fusion will be performed in a file 32 bits per channel (sometimes referred to as HDR container in shades 4294967296), that we will then "compress" (nothing to do with compression JPG) to fit in a 16-bit encoded file (a computer screen do not have sufficient dynamic range to display an HDR image): this is called tone mapping. Area by area, HDR fusion algorithms will choose the best exposed pixels in each shot, to compose an image in which all pixels are properly exposed.

Above all: know what you want to do with HDRI!

We will see that the HDR technique can produce very different from the same set of source images pictures. First of all, so you just need to know what you want to do!

There are roughly speaking two reasons that can motivate a photographer to make HDR:

Exceed the capacity of its digital camera to produce a photograph closer to the perceived vision and emotion felt by the human eye (sometimes called "photo-realistic" or TTHDR for "True Tone HDR")

Use the technique to create a surreal image typical "HDR Impressionist" or "texture extraction" close to a real painting or drawing

Here are two images produced from the same basic picture and the same picture without the use of HDR technique: