I've heard that human cone cells can detect only one color, so it's more like a Bayer sensor than a Foveon. So, if the human eye is close to the Bayer sensor, I wondered if color moire would occur in the human eye as in the attached photo. Will the human eye use a feature like a low pass filter like most cameras to solve this problem?

text text

bayer                     foveon

source : https://sudonull.com/post/184570-Hybrid-film-and-CCD-CMOS-matrix-is-not-possible-Try-saying-it-Foveon-X3

  • \$\begingroup\$ Please see Filter for RGB separation and its effect on the image for how both the cones in the human retina and the color-filtered photosites on Bayer masked sensors create color out of specific combinations of wavelengths of light. Your question makes some rather misinformed assumptions about both. Hint: the peak sensitivities for our S, M, and L cones are not R, G, and B. Neither are the peak transmissivity of the three colors used in a Bayer mask, all the cute little pictures of red, green, and blue squares on the internet notwithstanding. \$\endgroup\$
    – Michael C
    Feb 1, 2021 at 17:21
  • \$\begingroup\$ Neither our cones nor the photosites behind color filters of Bayer masks are only sensitive to one color, either. They both have a LOT of overlap in the wavelengths to which each is sensitive. \$\endgroup\$
    – Michael C
    Feb 1, 2021 at 17:34

2 Answers 2


It's more complicated than this because:

  1. Moiré appears when there is a slight difference of spatial frequency between the image and the sensor. But the cones and the rods, being organic, are not in a neatly equally spaced grid so you cannot define a spatial frequency for them (or at least there is a frequency range so the moiré is severely attenuated).

  2. Our eyes are in constant movement so the image isn't fixed on the retina and any moiré would be a very transient effect.

  • \$\begingroup\$ Re #2. Moving the camera while shooting video with a camera that has a Bayer filter, but no AA filter, does not hide moire. \$\endgroup\$
    – xiota
    Feb 1, 2021 at 16:47
  • \$\begingroup\$ Re #1. Autochrome has randomly distributed color filters. \$\endgroup\$
    – xiota
    Feb 1, 2021 at 16:49
  • \$\begingroup\$ @xiota Moving the camera does shift the shape and location of the moire, though. \$\endgroup\$
    – Michael C
    Feb 1, 2021 at 17:35
  • 1
    \$\begingroup\$ 3. The image is really created in the brain. The eyes are delivering visual information that the brain interprets to form a full image. So even if the eyes would see a moire pattern, the brain would probably ignore it. \$\endgroup\$
    – Pete
    Feb 2, 2021 at 7:52
  • \$\begingroup\$ Re #1, Fuji created the X-trans sensor (which is not a bayer based sensor), that has a layout that should minimise moire. In fact to such a degree that they don't put a low pass filter in front. \$\endgroup\$
    – Pete
    Feb 2, 2021 at 7:58

With regard to moire, not really. Both Bayer masked sensors and Foveon sensors have regular grids of photosites that contribute in a large way to moire. Cones (and rods, for that matter) in the human retina are not distributed in such a regular pattern. So in that respect, there's a fundamental difference between human retinas and any kind of digital camera sensor.

With regard to how the various wavelengths of light are interpreted to create colors, both Bayer masked sensors and, to a lesser extent, Foveon sensors mimic the way our brains use the overlapping response curves of the three types of cones in our retinas to create color out of combinations of various wavelengths of light.

There is no color implicit in light. Color is the result of the perception of light and is a construct of the brain (or image processor) that perceives (or records/interprets) certain wavelengths of light.

Neither our cones nor the photosites behind color filters of Bayer masks nor the three layers of a Foveon sensor are only sensitive to one color. They all have a LOT of overlap in the wavelengths to which each type is sensitive. Without that overlap, our brains could not create colors out of our perception of different combinations of various wavelengths of light.

enter image description here
Please note that the actual colors to which each type of retinal cone is most sensitive are not "Red", "Green", and "Blue"!

It is true that with a Bayer mask the light is filtered with either a Red, Green, or Blue filter¹ over each pixel well. But there's no hard cutoff where only green light gets through to a green filtered pixel or only red light gets through to a red filtered pixel. There's a lot of overlap. A lot of red light and some blue light gets through the green filter. A lot of green light and even a bit of blue light makes it through the red filter, and some red and green light is recorded by the pixels that are filtered with blue.

enter image description here
Please note that the peak transmissivity of each of the three Bayer mask filter colors is not "Red", Green", and "Blue", but rather closer to the same wavelengths of light to which our retinal cones are most sensitive. In a very real sense, our Bayer masks are more "Yellow - Lime - Violet" than they are "Red -Green - Blue".

In the same way, each layer of a Foveon sensor responds to more than only red, or only green, or only blue. The values recorded by each layer of a Foveon sensor are not directly translated to RGB values. Color channel multipliers still need to be applied to account for the differences in the peak sensitivity of each layer compared to the colors we use in our color reproduction systems. Just as there is not a one-to-one correspondence between the colors to which each of our cones is most sensitive and the colors our RGB or CMYK color reproduction systems use, there is no one-to-one correspondence between the colors to which each layer of a Foveon sensor is most sensitive and RGB. And that's before we even begin to account for the different types of light with various spectral distributions that are illuminating our subjects when we use those sensors.

¹ The actual colors of the Bayer mask in front of the sensors of most color digital cameras are: Blue - a slightly violet version of blue centered at 450 nanometers, Green - a slightly bluish version of green centered on about 540 nanometers, and Red - a slightly orange version of yellow. What we call "red" is the color we perceive for light at about 640 nanometers in wavelength. The "red" filters on most Bayer arrays allow the most light through at somewhere around 590-600 nanometers. The overlap between the "green" and "red" cones in the human retina are even closer than that, with "red" centered at about 565 nanometers, which is what we perceive as yellow-green.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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