|
Digital week, day 4 : What's the problem with digital... |
|
|
|
Thursday, 13 January 2005 |
|
It all comes down to physics : a CCD sensor's sensitivity and dynamic range is linked to the size of each pixel. It's the same thing with film, sensitivity is directly linked to silver grain size.
Now. Sensor size is a very important parameter in digital camera, and most digital cameras have small sensors, and the sensor size has not changed in the last few years, only pixel number as increased, and the only logical way to fit more and more pixels on a same-size sensor is to make the pixels smaller...
CCD sensors are manufactured from silicon wafers, and those wafers have defects, which means that if a ccd chip contains a defect, it has to be rejected. CCDs are difficult and costly to make in larger sizes, because the larger the size, the higher the number of CCDs with problems that have to be rejected. That's why sensor cost is not a function of area only. The way to increase the dynamic range (which is the difference in illumination intensity between what the sensor "sees" as white and what the sensor "sees" as black) the way is to increase the size of pixels, and increase the size of the sensor. That's what is done in Medium Format digital back, and that's why they cost as much as they do (in the 10 000$ range...)
One solution would be to make a 2 or 3 megapixels camera with CCDs of the same size as current 5, 6 or 7 megapixels cameras, but with individual pixels being much larger. Fuji has a technology that uses this idea partly, by using two sensors for each pixel, one smaller that can read "high" lights, and one larger that reads "low" lights.
The digital world is falling victim to the megapixels race, I think photographers would be better served if quality was favored over quantity... |
