Working closely with Sony, Phase One introduces IQ3 101MP Trichromatic medium format digital back. It uses "a new CMOS sensor and Bayer Filter color technology, available only through Phase One, we have given the photographer 101-megapixels of creative possibility in never-before possible color definition." It is said to be capable of replicating, closer than ever, the color definition that the human eye sees.
"Designed around the concept of mimicking the dynamic color response of the human eye, we have physically customized the Color Bayer Filter on the 101-megapixel sensor to tailor the color response. This allows the Digital Back to capture color in a new way, unlike anything else.
The Phase One Trichromatic Philosophy is a promise that where color and quality can expand, while others may be satisfied with what they have, Phase One will always strive for perfection."
There is not much more info released about the new image sensor:
A - All prism cameras have this color separation characteristic - it is almost unavoidable.
ReplyDeleteB - filter companies have been selling filters for years to put in front of Bayer cameras to provide better channel separation, but, unfortunately...
C - lack of channel overlap renders the sensor unable to discriminate changes in wavelength that occur inside any channel band. This effect is familiar to anyone using broadcast (prism) cameras to shoot concerts - the purple lights look blue on the video display.
In no way does high channel separation mimic the human eye. The red and green channels overlap about 80% in the eye. The blue overlaps green less but the overlap is still substantial.
Perhaps they should publish the DeltaE numbers for this sensor to support their assertions. The pictures on their website look postery for a reason.
I would add to C that either they have to translate into a sRGB or similar color space which is based on spectral overlapping function similar to the human observer. This color space translation will contain increased off-diagonal coeffcients compared to standard Bayer dye filter which decrease finally the SNR. It is the same effect as for the Foveon where in opposite the overlap is higher than the target colorspace. The bad thing in comparison to Foveon is that the light is filtered before being converted into electrons.
DeleteSo in conclusion the "mismatched" spectral overlap to the final colorspace created SNR reduction through off-diagonal coeffcients w/o taking advantage of it.
Please P1 give me a clue about your claims!
I wonder if someone read my comment almost two years ago: http://image-sensors-world.blogspot.ca/2015/10/new-cfa-said-to-have-better-color.html - not a copy of my suggestion but an alteration of the Bayer as I mentioned would be one part of the chain towards better color (also need better output device and intervening processing, which Phase One has).
ReplyDeleteTwo for two today?