In many articles about CMOS sensor, I usually find that the relationship between readout time and frame rate is:

(1) Readout time=1/Frame rate

But Jia, C., & Evans, B. L. (2013, December) says that the readout time is usually about 60% to 90% of the time interval between frames. It means

(2) Readout time=0.6/Frame rate to 0.9/Frame rate

Which information is correct? If (1) is correct, what is the effect of the 'reset time' of each row? Is it that the 'reset time' is too small compared to the 'read time' of each row so that the 'reset time' is ignored? If (2) is correct, what are those 10% to 40% of time for?

  • \$\begingroup\$ Your question could probably use a little more context. What type of camera are you talking about? Mirrored or mirrorless? Mechanical shutter or electronic shutter? Etc. \$\endgroup\$
    – Michael C
    Commented Jul 14, 2015 at 22:28
  • \$\begingroup\$ It is electronic shutter. I'm not talking about any specific camera, but a general camera using CMOS sensor with electronic rolling shutter. (1) can be found in many articles explaining CMOS sensor with electronic rolling shutter. But let's say it is mirroless, what is the answer for my question? \$\endgroup\$
    – opmfan
    Commented Jul 15, 2015 at 2:53

1 Answer 1


As far as I can tell, the sensor has a fixed readout time. When using a frame rate slightly slower than the readout time, the sensor momentarily sits idle at the end of each readout, before beginning the next one - rather than slowing down the readout, which the sensor can't do.

I don't know why they give that specific range (60% to 90%) but maybe they're talking about a very specific situation rather than a general sense. I didn't read the whole paper.


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.