|
|
Noise
|
| The Cause: Sensor Noise
Each pixel in a camera sensor contains one or
more light sensitive photodiodes which convert the incoming light
(photons) into an electrical signal which is processed into the
color value of the pixel in the final image. If the same pixel would
be exposed several times by the same amount of light, the resulting
color values would not be identical but have small statistical
variations, called "noise". Even without incoming light, the
electrical activity of the sensor itself will generate some signal,
the equivalent of the background hiss of audio equipment which is
switched on without playing any music. This additional signal is
"noisy" because it varies per pixel (and over time) and increases
with the temperature, and will add to the overall image noise. It is
called the "noise floor". The output of a pixel has to be larger
than the noise floor in order to be significant (i.e. to be
distinguishable from noise).
|
| |
| The Effect: Image Noise
Noise in digital images is most
visible in uniform surfaces (such as blue skies and shadows) as
monochromatic grain, similar to film grain (luminance noise) and/or as
colored waves (color noise). As mentioned earlier, noise increases with
temperature. It also increases with sensitivity, especially the color
noise in digital compact cameras (example D below). Noise also increases
as pixel size decreases, which is why digital compact cameras generate
much noisier images than digital SLRs. Professional grade cameras with
higher quality components and more powerful processors that allow for
more advanced noise removal algorithms display virtually no noise,
especially at lower sensitivities. Noise is typically more visible in
the red and blue channels than in the green channel. This is why the
unmagnified red channel crops in the examples below are better at
illustrating the differences in noise levels. |
| |
|
Blue Sky Crop |
A |
B |
C |
D |
E |
|
RGB |
 |
 |
 |
 |
 |
|
Red Channel |
 |
 |
 |
 |
 |
|
Camera Grade |
Professional |
Prosumer |
Prosumer |
Prosumer |
Crop C after 123di noise reduction. |
|
Camera Type |
SLR |
SLR |
Compact |
Compact |
|
Pixel Size |
Large |
Large |
Small |
Small |
|
ISO |
100 |
200 |
100 |
800 |
|
Red Ch. St. Dev. |
1.8 |
2.5 |
5.6 |
22.6 |
1.4 |
|
| |
|
The standard deviation measured in a uniform
area of an image (in the above examples measured in the red channel)
is a good way to quantify image noise as it is an indication of how
much the pixels in that area differ from the average pixel value in
that area. The standard deviation in the noisy examples C and D is
much larger than A, B, and E. Crop E shows that noise reduction can
go a long way.
|
| |
Long Exposure "Stuck Pixels" Noise
|
Another type of noise, often referred
to as "stuck pixels" or "hot pixels" noise, occurs with long
exposures (1-2 seconds or more) and appears as a pattern of
colored dots (slightly larger than a single pixel). As
explained in the noise reduction topic, long exposure noise
is much less visible in the latest digital cameras.
|
 |
|
|
