This discussion looks into the differences between two sensor technologies which drive modern day digital cameras, CCD and CMOS. The digital sensor is what the light coming through the lens hits inside the camera. Both CCD and CMOS these sensors essentially use the same concepts of photoelectric to convert electrons from light. The sensor collects the light, processes it and then converts it into electrons. These are then transmitted to the image processing engine via internal wiring to produce the final images that we see on the LCD monitor. You may have already seen an image of a digital sensor on the internet. Heck, they are everywhere these days!
From smartphones to security cameras to state of the art digital SLRs, either of these two technologies are at the heart of a majority of the digital cameras out there. Learn more about how to use DSLRs with this course.
So, how they look? Well, if you own a DSLR camera you have probably the easiest option to see them. Simply un-mount the lens from your camera and flip the mirror up. What you see gleaming inside is the sensor. Depending on the model and make it could be a CMOS or a CCD. But if it is a new model then there is a high chance it is a CMOS sensor. For more details on digital sensor technology you may refer to these course on Udemy.com
CCD vs CMOS – Light Gathering and Conversion to Electrons
The argument as to which one is the better technology has been razing on for some time now. There are arguments both in favor of and against of both these technologies. Before the introduction of CMOS technology, digital cameras were all powered by CCD sensors. CCD stands for Charge Coupled Device and they are the more power hungry of the two types. CCD sensors use a system where the actual processing of the charge obtained by each pixel is carried out off-pixel. So, the amplification of the signal, conversion to electrons and then transfer to the image processing is all outsourced. The result is that the area occupied by each pixel is wholly dedicated to the capture of light.
The CMOS sensor on the other hand uses a different algorithm for converting light to charge. The entire process is handled by the pixels themselves. This means right from the capturing of light to conversion into charge and amplification of the signal, everything is handled by the individual pixels themselves. All these complex circuitry and conversion metrics reduce the area in each pixel that is dedicated to collecting light. As a result, a lot of the light strikes these small transistors instead of falling on the light sensitive photodiodes. The result is higher noise to signal ratio in CMOS sensors. This is why CMOS sensors have been found to be inferior to CCD sensors in very low light conditions, like you can learn about in this course. This is evident when you use a high ISO number on a CCD powered camera compared to a CMOS powered one.
For the basic architectural reasons, output of the CCD sensors are analog while that of the CMOS sensors are digital. CMOS sensors have their own charge to voltage conversion system, signal amplification and noise elimination system. These allows camera manufacturers to basically design a digital camera on a chip. CCD based camera require additional mechanisms to convert the signals to electron, then correct for noise as well as convert the voltage to digital image.
Uniformity of the Signal Output
Since each pixel of a CMOS sensor has its own charge to digital bit conversion system, maintaining an uniformity in as much as image quality and exposure is difficult in different (but uniform for all pixels) lighting conditions. CCD’s are much better off in this area. For a period of time CCD’s were thus preferred as they offered better response and handled different lighting conditions uniformly across all pixels. However, off late, better conversion algorithms in CMOS systems have started to close this gap and performance too has been comparable.
The same architecture, however, had an advantage when it came to correcting overexposure (in certain cases) affecting a small percentage of the sensors, without affecting the overall picture. Meaning, if a few of the pixels suffered from overexposure that can be corrected easily on a CMOS sensor (because it handles signal conversion for each pixel individually) than on a CCD sensor (which is dependent on conversion off-sensor.
Speed of Conversion
CMOS sensors, in spite of their less than satisfactory results in lowlight conditions, are found to be faster in terms of image processing when compared to CCD. Since each pixel does its own conversion from analog to digital signal, they work independent to each other. Additionally, the transfer of signals to the image processing engines is also done using their independent wiring, making the whole process faster. CCD sensors on the other hand have to depend off-sensor in order to convert the light signals to voltage and then transferred to the image processing engine. Each individual pixel propagates its signal to the next one and this continues till the pixel closest to the output node is reached where it is offloaded and sent to the processing engine for conversion to digital image. This results in a much slower transfer rate.
It is interesting to note that both CCD and CMOS sensors were developed roundabout the same time (1960’s). However, the available manufacturing technology at that time meant that it was the CCD sensors that could be more effectively produced. Thus they were produced in greater numbers. In fact CCD sensors dominated the digital sensor market up until the 1990’s when CMOS sensor technology came back into reckoning due to improved manufacturing process.
One of the major reason why CMOS sensors are now being preferred over their CCD cousins is that they are less power hungry. Additionally, the growth in demand of on-chip camera systems which are cheaper to make and are comparable in performance to CCD powered systems is also an attributing reason. With the growth in sales of smartphones and iPhones (learn more with this course) there was need for an efficient and less power hungry sensor that could power the cameras on these phones. The industry demanded a sensor that is capable of being built into smaller forms. Thus CMOS systems came back into reckoning. Major investments were made to make it possible to manufacture CMOS sensors cost effectively leading in their wide-spread use.