As I understand the way ISO boosting works in a digital sensor, is that the voltage coming from each pixel is boosted (or suppressed) by an analog amplifier. So, for example, if the native sensitivity of the sensor is 100 ISO and the photographer asks for 400 ISO, then the voltage is amplified 4X before being sent to the ADC.

So, my question is whether there is one such amplification circuit for every pixel, or is there only one circuit and the output from each pixel is sent to the circuit in turn, or middle-of-the-road, is there a bank of amplifiers and pixels are sent in rows or columns to the amplification bank?

  • 1
    As the answers to the suggested duplicate say, the answer to the last paragraph is pretty much, "all of the above." It depends on the specific camera in question. – Michael C Aug 17 '17 at 4:54

The CCD imaging chip sports photosites in a grid pattern. The photosite contains a photodiode and a storage area. Hits by photons during the exposure generate an electrical charge. The more hits, the higher the charge. When the exposure is complete the charge is moved into the site’s storage area. Next each photosite transfers its charge into a transfer register. One by one the photosite charges are read out and converted to a voltage and each is then amplified. This reading and amplification happens row by row. It is this coupling of the data from one row to the next that gives this chip its name “charge coupling device. The CCD consumes more power because it converts photon hits to a charge and amplifies. The CCD requires additional processing functions as compared to a CMOS. Thus the CCD is slower.

The CMOS imaging chip also features row upon row of photosites arrange in a matrix. Each site contains a photodiode that does coverts the photon hit to a charge. The charge is not stored in the photosite; it is moved to an adjacent part of the chip or passed to adjacent chip. Now it is changed from a charge to a voltage and amplified. This design allows each photo site to be handled individually rather than row by row. With a CMOS most of the processing functions are handled independent of the imaging chip. CMOS stands for complementary metal oxide semiconductor.

  • I've downvoted - not in any way because this isn't a good answer, but because it's just a good answer to a completely different question. :) – James Snell Aug 16 '17 at 22:36

Not the answer you're looking for? Browse other questions tagged or ask your own question.