1
\$\begingroup\$

Newer Samsung smartphones have sensors with absurd pixel counts like 108 Mpx where pixel size is 0.8 micrometers but they are grouped in 3x3 arrays of the same color making them effectively 2.4 micrometers long with 12 Mpx pixel count.

I strongly suspect that the main reason for that is marketing since they can sell "more pixels" in their cameras and 12 Mpx camera with bigger pixel size wouldn't sound as flashy.

Are there any reasons why the technology should be better than an equivalent sensor with larger pixels with the same area?

I could argue that it could be worse since each pixel probably has some edges which makes the receptive area smaller compared to a larger pixel.

\$\endgroup\$
1
  • \$\begingroup\$ I can't be sure about it because I do not have deep knowledge of how CMOS sensors work but if you check DXOmark.com scores among same sensor sizes you will see that best low light scores ("Sports") are from cameras which do not have the smallest pixels. \$\endgroup\$ Oct 24, 2023 at 15:32

1 Answer 1

1
\$\begingroup\$

I could imagine 2 potential advantages:

  • pixel binning could allow a better night mode (for example, by keeping the colors which are most represented in the pixels of a group),
  • electronic 2x or 3x zoom could be optically better because instead of interpolating pixels, the processor can simply use all the pixels.

I have a Samsung Galaxy A51 with a 48 megapixels sensor, generating 12 MPx images. FWIW, here are the results of my personal tests:

  • I would have liked a better night mode, but obviously Samsung did not (could not?) do much better than usual.
  • I tested the 2x zoom on some foliage (lots of details and lots of shades of colors). I took 2 pictures, one à 1x, and one at 2x. I am convinced there are details in the 2x picture that could not have been extracted from the 1x picture. IOW, it seems that at 2x, the A51 does not use pixel binning but instead uses each of all the center pixels. It would be interesting if someone did the same test on a nonapixel device at 3x zoom.

Edit: as @FarO commented, sensor resolution is only useful if the lens lets enough details go through. Indeed, the only details which appeared in my 2X tests were those with some level of contrast; this may explain at least partially why the colors on the leaves seemed a little strange.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ This answer is correct. For info: the A51 will get better details only if you have plenty of light. If it's not a very bright environment, the much higher noise will eat any advantage. Also, in general light diffraction limits (= severely reduces) details on the sensor based on the aperture value: a f/2.8 lens will not provide any reasonable detail below 2.8 microns. Smartphones have usually f/1.8 with pixels of 0.8 um, so "un"binning can give some more details ONLY when the details involve a sharp contrast. Softer details are NOT there. You may want to add this to the answer. \$\endgroup\$
    – FarO
    Oct 23, 2023 at 20:11
  • \$\begingroup\$ Comments have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on Photography Meta, or in Photography Chat. Comments continuing discussion may be removed. \$\endgroup\$
    – scottbb
    Nov 11, 2023 at 21:24
  • \$\begingroup\$ All three colors of every pixel are interpolated in a digital image. The colors of CFA's (blue-violet, slightly yellow green, and yellow-orange) do not match the colors used by our screens (RGB) nor printers (CMYK). Removing the incorrect information in the second bullet point would therefore improve the answer. \$\endgroup\$
    – Michael C
    Nov 15, 2023 at 3:23

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.