Most photographers who prefer rendering from CCD sensor assemblies are not imagining things. However CCD sensor assemblies are not inherently superior to CMOS assemblies in therms of color rendering or monochrome tonality.
The most important aspect of color rendering in-camera are the passive devices in front of the sensor bed. The characteristics of the IR filter (if there is one) and the color-filter array assembly are important and vary from brand to brand and between models for the same brand.
The preference for CCD rendering is actually a preference for the camera's color-filter array assembly
The Bayer reconstruction algorithms assume only red, blue or green light frequencies are present for a specific sensor site. However
all filters are imperfect. It is more expensive to manufacture color-filter arrays that minimize light from unwanted frequencies. That is, the red sites will be contaminated with some amount of blue or green light as are the blue and green sites. There are coefficients in post-production rendering programs that correct for filter imperfections. This is one reason why raw from different cameras use different rendering profiles. However nothing beats collecting high-quality data in the first place.
Both CMOS and CCD sensors use the photoelectric effect to generate analog signals (convert light energy to electrical charge). The only difference is the CCD technology requires separate chips to complete the data stream output while CMOS technologies integrate all of the data stream electronics within the main sensor chip. The later technologies results in less electronic noise (read noise). This is how come CMOS cameras enjoy a signal-to-noise ratio and analog dynamic range advantage compare to CCD cameras.
Both CCD and CMOS employ pinned photodiodes to utilize the photoelectric effect. The output of both CCD and CMOS photo sites are analog DC voltages, which are the signals. DC voltages alone can not have any impact on color rendition whatsoever. Instead they are just an estimate of the light energy a particular pixel collects (and the lower the signal-to-noise ratio, the less certain the estimate becomes). Other factors such as the transmission properties of the color filter array, the IR filter transmission properties and even the frequency response of the PIN diodes themselves could affect the DC voltage amplitudes and could impact Bayer image reconstruction. But these variables will have the
same impact for both CCD vs CMOS.
If you think I am misrepresenting the CCD vs CMOS story please take the time to read these links.
This is a fairly non-technical general comparison of CCD and CMOS sensor technologies.
Here are is a detailed technical review article from a peer-review journal.
link 1
About midway through this
presentation the same author shows
data comparing the quantum efficiencies of CMOS vs CCD sensors. QE determines the sensors sensitivity to light. It affects the signal component of the signal-to-noise ratio.
This
presentation summarizes the history of image sensor development and explains in detail the physics and engineering behind the superior signal performance (
which is not color performance) of CMOS devices.