Originally posted by enricob Does this mean that a 50 MP 33x44 sensor camera should be less sensitive to motion than a 50 mp 24x36 sensor camera?
Does this mean that a 24 MP 24x36 sensor camera should be less sensitive to motion than a 50 mp 24x36 sensor camera?
Did I get it right?
If I get it right, a higher resolution (50mp) camera as 645z is more sensitive to motion and therefore forces shorter shutter speed and upper ISO values. Someone suggests a lower resolution (24mp) camera in case you need shooting with longer shutter speed and lower ISO values because it's less sensitive to motion. Do you agree? What would you prefer?
Thank you,
Enrico
First, it must be emphasized that we are addressing angular motion, and not translation motion, which would have an effect in the near field. (We have to note that it is pixel size vs. focal length that determines image translation due to angular motion.) When you use the word "more sensitive" I interpret it to mean blur extending farther over across some number of pixels. Fundamentally, we are addressing "smear" and
any motion will cause smear; the only question is how noticeable it is. In principle, one could take the image from a 50 Mp camera and derez it to a 25 Mp equivalent (to the extent that a linear root 2 derez was achievable) and the blur would be less noticeable even as the pixel resolution would be more noticeable (in a large enough print).
Second, we must keep in mind that pixel size affects the noise level that a given ISO value will yield in that camera, so in reality, smear sensitivity and noise sensitivity may be interactive in a given result. Other ways of looking at pixel size effects include:
- From the point of view of modulation transfer function effects, the product of all the MTFs to yield the MTF of the actual read out image is being compared to the same final MTF times the MTF of a smear motion. The effect will be a reduction of the higher frequencies in the image, but if the pixels are large enough in angle space the effect will be initially hidden for small smears.
- As another attempt at description, assume a diffraction free optical system, and infinitely fine grain film. The image will have infinite resolution. Then read it out by sampling the density pattern over a series of contiguous areas representing pixels. Compare to reading out a smeared version of this image. I can't claim that larger pixels provide an improvement over smaller pixels when the end result is observed with some level of degraded image resolution (eye, monitor, etc.). No matter what the pixel size is, smear always spills the image across more than one pixel, but not necessarily over exactly two pixels.
tl;dr: I think you got it right, given a
lot of assumptions (that if ignored may lead to misleading understanding). I think I prefer more resolution because there is more information to work with such that post processing may be able to be more effective (for someone skilled in the art).