Subject resolution is the key to how much an image can be usefully cropped - ie how many pixels are covered by the desired bit of the image. A number of factors will interplay here.
Pre-sensor - focal length, ie how big the image is projected on the sensor. Also, most crucially, subject distance from the camera.
Sensor - pixel density rather than pixel dimensions of the whole sensor (assuming that the subject does not use the majority of the sensor). Also, although not strictly relevent to the current discussion is pixel quality - I remember a debate about small vs large pixels, the upshot of which was that larger pixels had better dynamic range and less noise generaton than small pixels.
Post-sensor - interpolation. Many approaches to this, which I am not well enough informed about to debate, but the fundamental priciples are less interpolation introduces fewer artifacts, and a good quality starting point will give a better result than a poor starting point. (Seems obvious, but this is where the pixel quality debate comes in)
Finally output - I dont need 16Mp to produce a web image, but I do need it for a decent A3+ print, or for picture libraries.
To be really objective and be able to compare things across different platforms, from APS to 5"X4" scanned trannies via digiscoped images, what we really ought to use is 'field of view', but I certainly cant visualised what the numbers mean as easily as I can visualise what the (35mm equivalent) focal length implies - daft really when you think about it, as the FOV is exactly what it says, but I think most photographers use the same shorthand as me.
Pre-sensor - focal length, ie how big the image is projected on the sensor. Also, most crucially, subject distance from the camera.
Sensor - pixel density rather than pixel dimensions of the whole sensor (assuming that the subject does not use the majority of the sensor). Also, although not strictly relevent to the current discussion is pixel quality - I remember a debate about small vs large pixels, the upshot of which was that larger pixels had better dynamic range and less noise generaton than small pixels.
Post-sensor - interpolation. Many approaches to this, which I am not well enough informed about to debate, but the fundamental priciples are less interpolation introduces fewer artifacts, and a good quality starting point will give a better result than a poor starting point. (Seems obvious, but this is where the pixel quality debate comes in)
Finally output - I dont need 16Mp to produce a web image, but I do need it for a decent A3+ print, or for picture libraries.
To be really objective and be able to compare things across different platforms, from APS to 5"X4" scanned trannies via digiscoped images, what we really ought to use is 'field of view', but I certainly cant visualised what the numbers mean as easily as I can visualise what the (35mm equivalent) focal length implies - daft really when you think about it, as the FOV is exactly what it says, but I think most photographers use the same shorthand as me.