1
\$\begingroup\$

I have been testing two different sensors, using 3 lenses between each of the sensors. I have also kept the exposure to 6ms between each sensor. I noticed a difference in measured DN pixel values between targets for each respective sensor, at the same exposures.

The sensor with a lower spatial resolution has higher DN values of targets. And, the higher spatial resolution sensor consistently has lower values.

Does spatial resolution at all affect the exposure of an image, and thus the measured pixel values? As in, perhaps photons are more distributed amongst more pixels in higher spatial res. images, resulting in lower DN?

\$\endgroup\$
6
  • 1
    \$\begingroup\$ Have you controlled for other variables, such as sensor temperature, base ISO, fill factor, front vs. rear illuminated sensor, etc.? Also, are the sensors made by the same manufacturer? \$\endgroup\$
    – scottbb
    Commented Dec 16, 2016 at 2:17
  • 2
    \$\begingroup\$ What term are you abbreviating as DN? \$\endgroup\$
    – Michael C
    Commented Dec 16, 2016 at 2:44
  • \$\begingroup\$ I guess Dark Noise? \$\endgroup\$
    – mattdm
    Commented Dec 16, 2016 at 3:16
  • \$\begingroup\$ IIRC, DN stands for "data number" or "digital number," and it refers to the pixel values as read straight from the sensor. \$\endgroup\$
    – Caleb
    Commented Dec 16, 2016 at 4:01
  • 1
    \$\begingroup\$ @Alexander What is the purpose of your test? What problem are you trying to solve? How are you hoping to apply what you learn from the tests? \$\endgroup\$
    – Michael C
    Commented Dec 16, 2016 at 18:44

1 Answer 1

2
\$\begingroup\$

It all depends upon how one defines exposure. If one defines exposure as the total number of photons captured per pixel then pixel density will most definitely affect exposure.

But in photography exposure is normally expressed in terms of field density. That is, in terms of photons per area unit of the film or sensor's surface. When the raw sensor data is used to create a viewable image the number of photons measured per pixel is normalized to account for the varying sizes of pixel wells (more properly called sensels) in much the same way that the brightness of prints of various sizes from the same negative are normalized so that a print with twice the surface area of another print from the same negative does not wind up being half as bright.

\$\endgroup\$
1
  • \$\begingroup\$ I think this answers my question Michael, in that I will need to get more specific information from the manufacturers regarding the specifics of their sensors. From just viewing the imagery it is hard to determine how much welling and normalization differs between sensors. \$\endgroup\$
    – Alexander
    Commented Dec 16, 2016 at 16:05

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.