University of South Carolina

PHYSICS AND THE VISUAL ARTS
Notes on Lesson 12

Things to remember from Lesson 12.


We saw earlier that the human visual system has a behavior we called color constancy. By that we meant that a green shirt looks green whether we see it in outdoor sunlight or indoors under flourescent lights or incandescent lights. As long as the light can be considered as "white" light, we recognize the colors of objects. There may be small variations and we have seen some of these with metamers. However, our vision works remarkably well regardless of the light source.

Photographic films do not behave like our eyes. A film that is designed to produce good colors with sunlight will not do so with incandescent lighting or with fluorescent lighting. The reason is that incandescent light is deficient in blue compared with sunlight and so film used with incandescent light needs to have a blue sensitive layer that has a higher sensitivity than needed in sunlight. You may want to review the examples made with the Polaroid camera shown in the lecture.

Color filters are used with cameras to compensate for the light and to balance it with the choice of film. Indoor film (often called tungsten film) is typically balanced for light with a color temperature of 3200 K. If such film is used and the light is from household lamps (at around 2900 K) everything will be too red. If the same film is used outdoors everything will be much too blue. An orange filter can reduce the blue that gets to the film when it is used outdoors.

What color filter should be used if you need to take outdoor film indoors?

In recent years the sale of electronic digital still cameras (DSC) has exceeded the sale of film cameras. These DSCs use no photographic film. Instead they contain a photosensitive device that consists of a huge array of tiny photodetectors. The most common form of photodetector is the CCD or charge-coupled device. The smallest element of a picture is called a pixel, (short for picture element). The number of individual photodetectors in the CCD determines the number of pixels available for generating the image. In most cameras a filter mask is placed over the array so that the individual photodetectors carry information about only one of the primary colors: R, G, or B. In most cameras the filter mask is designed so that each square of four detectors has one filtered red, one filtered blue, and two filtered green. This particular choice for filters is known as a Bayer mask.

The resolution in the CCD detector depends on the overall size of the array and how many individual photodetectors make up that array. Each year it seems that the camera manufacturers increase the number of pixels for their cameras. The minimum number of pixels needed to make small (4" x 6") prints from a DSC is two million. Cameras with fewer pixels can be used to make photos for use on the internet. If you want to make larger photos, then you should use a camera with more pixels. Of course, they cost more, especially if you go to the extremes of 11 million pixels or more.

Most of the CCD cameras now in use a 4 x 3 format. That is, the pictures have a rectangular form that is four units wide by three units high. That is the shape of conventional television (before HDTV). This is slightly narrower than the format of 35-mm cameras whose images have a 3 x 2 shape. Some of the newer DSCs have large photodetector arrays with the same size and shape as the image size on 35-mm film. These are the cameras with the most pixels and costs. Most of the earlier DSCs have small image areas compared to that of 35-mm film. Consequently, they use smaller lenses.

Many of the newer cameras use a different chip, one known as a CMOS chip. The CMOS technology is the standard computer chip technology that has been around for years. However, when DSCs were being first developed, the CCD technology gave better images than the CMOS detectors of that day. By now the CMOS image detectors have been developed to stage where they give images comparable to or better than those made with CCD chips. Because of savings that can be made in manufacturing and chip design, CMOS detectors will likely supplant the CCDs in the long run. The CMOS chips are simpler and require fewer external supporting devices for their operation.

One feature of most DSCs is that they have built-in color correcting ability. So not only do the automatically control for focus and exposure, the also adjust to the type of lighting. The relative RGB signals are adjusted to that white is white regardless of the color temperature of the source light. See the exampled in the lecture demo.

Some of the digital cameras use the EV system to control light. The exposure value or EV system was developed in the 1950s. The EVs are given in whole numbers such as 1, 2, 3, 4, . . .. Each increment of 1 EV reduces the exposure by 50%. The definition of EV is

EV = log2[(f-number)2 x ISO/(shutter speed x 100)].

Thus the exposure value includes the effect of aperture, shutter speed, and film speed.

Links to other sites.
Kodak primer on digital cameras.
This link for CMOS sensors and other DSC info.
Micron's site for CMOS advantages.


Back to Physics 153 Home.
Back to Physics Dept web page.
Back to USC.

Maintained by rjones@mail.psc.sc.edu
Last Modified: 10/21/04