What is the dynamic range of a RAW file?

0
0
0
s2smodern

 

The dynamic range of the naked eye is large (approx. 12 stops). On a sunny day we'll not only see the sun-lit landscape, but at the same time we are able to distinguish what is happening in the shadows.
Our eyes really are able to bridge an enormous light intensity difference.

We are able to see dim stars in the sky and at the same time the houses under bright city lights. This is almost double the dynamic range of  slide film (6 stops). Digital camera and regular analog films - depending on brand, type, ISO range and noise - have a usable dynamic range between 6 and 9 stops. Will a RAW file deliver you a larger dynamic range?

dynamisch bereik raw, raw lineair, raw bestand

Overexposure and Underexposure simultaneously

A skier wearing a dark suit reflects hundreds of thousands times less light than the white snow around him. Nevertheless the human eye is able to see both the detail in the dark suit and in the snow at the same time. Most cameras can't. A camera with a low dynamic range will turn the snow into a white overexposed stain without any detail. All details in the dark suit will melt together to a single black blot. When the dynamic range of your camera is unsufficient for this scene, you will find both overexposure and underexposure in the same image.
In Photoshop you can use the shadow / highlight function to correct under- or overexposed ranges in a picture. HDR software is even more powerfull. However, both detail in a completely black (0,0,0) or completely white (255, 255, 255) spot are lost forever.

An eye can bridge a larger dynamic range than a camera. But that's not all. It even gets worse. LCD screen, printer or printing service all have an even smaller dynamic range than a camera. That's why a picture made at a sunny day might disappoint you, because you don't recognize all the details in shadows and highlights you actually saw when you took the picture. In the example of the skier, you have to choose as a photographer:

- either well exposed snow, combined with a completely  underexposed  silhouet of the skier,

- or heavily overexposed snow, combined with a well lit skier

- utilize exposure bracketing and HDR (don't forget to ask the skier to stop moving during the bracketing series, for the best results)

Total dynamic range versus usable dynamic range

The larger the dynamic range of a camera, the less stringent a photographer has to choose his exposure settings. The risk of either over-exposed and / or heavily under-exposed parts in your pictures will diminish, retaining detail in both shadows and highlights. In our camera reviews, CameraStuffReview distinguishes between total dynamic range and usable dynamic range of a sensor (see: How we test cameras). The total dynamic range of a camera of a good contemporary camera almost equals the dynamic range of the human eye. However, the signal to noise ratio in the shadows is low, even at low ISO settings. The signal to noise ratio in the dark tones is so low, that if you lighten the shadows while editing your pictures, the picture will not be suitable for a print. The usable dynamic range lies between 4 to 6 stops lower than the total dynamic range. In practice, it's the usable dynamic range which is important for a photographer, especially for an HDR photographer.

RAW or jpg?

In our reviews we choose to develop RAW files without noise reduction. As a result, it might seem as if the usable dynamic range for RAW files is less than that of jpg files. Of course this isn't true. Our tests show that it is possible to realize the same total dynamic range for RAW and jpg files. The usable dynamic range of RAW files  - after careful development in a good RAW converter - is at least equal, but usually better than the usable dynamic range of a jpg file.

Dynamic range vs. bit depth

Dynamic range is something different than bit depth / 8 bits and 16 bits data storage. Dynamic range tells you the number of stops between black and white. Ideally, the bit depth gives you the number of grays between black and white that a camera can distinguish. A digital camera with a dynamic range of 6 stops can store an image in 256 levels of red, green and blue (8 bits) or 16.384 levels for each RGNB channel (14 bits). For the two camera's in this example, both with a dynamic range of 6 stops, the dynamic range doesn't change if you choose for a 14 to 8 bits sensor; overexposed remains overexposed and underexposed remains underexposed. But the 14 bit Raw file will show much more color gradations. See also RAW or jpg: posterisation.

RAW

Number of levels for each zone

Zone / number of stops:

Fraction in zone

8 bit
(256 levels)

10 bit
(1024 levels)

12 bit
(4096 levels)

14 bit
(17712 levels)

1

1/2

128

512

2048

8192

2

1/4

64

256

1024

4096

3

1/8

32

128

512

2048

4

1/16

16

64

256

1024

5

1/32

8

32

128

512

6

1/64

4

16

64

256

7

1/128

2

8

32

128

8

1/256

1

4

16

64

9

1/512

-

2

8

32

10

1/1024

-

1

4

16

11

1/2048

-

--

2

8

Sensor versus the human eye; linear vs. logarithmic

Most CMOS sensors  behave linear: their signal will be twice as high if the amount of light which reaches the sensor doubles. Our eyes behave differently: we experience 10 times as much light as twice as bright. Digital cameras contain 10, 12 or 14 bit CMOS sensors. They can describe 1024 levels (10 bit), 4096 levels (12 bits) or 17712 levels (14 bits) , respectively. In the table above you can see what this means in terms of the zone system. The zone system consists of 10 or 11 zones, where each following zone is half as bright (1 stop) as the previous zone. The maximum signal for a 10 bits sensor is 1024 (max of zone 1). Divide the amount of light in half and the sensor signal will reach 512 maximum (max of zone 2). A 10 bit sensor has so little signal left for the dark zones, that the signal to noise ratio will become so bad, that the image is no further usable for a beautiful, noise free picture. A camera with a 10 bit sensor will have a usable dynamic range of approximately 5 stops. A (theoretical) 8 bits sensor has such a low usable dynamic range, that it becomes useless for photography.
But, I hear you think, how is it possible to obtain a high quality image out of a 8 bits jpg file? Camera and computer use a trick, called gamma, of which we'll go more into detail in a forthcoming episode.

Ivo Freriks
Author: Ivo Freriks
With Camera Review Stuff I hope to make a modest contribution to the pleasure that you get from photography. By testing cameras and lenses in the same way, evluating the results and weighing up the pros and cons, I hope to help you find the right camera or lens.

0
0
0
s2smodern

Leave your comments

Posting comment as a guest.
Attachments (0 / 3)
Share Your Location