gdanmitchell Offline Upload & Sell: Off
|
anorphirith wrote:
50mpx is good only if the lenses you mount on the camera can actually resolve it all.
Note exactly true.
It is certainly true that if your goal in using a higher resolution sensor is to make larger prints you'll want to use excellent lenses. (Judging from some of the sample shots that Canon released along with the new announcements, there are a number of lenses that work well with 50MP images.)
In addition, the "boundary" between being able to resolve and not being able to resolve is not a binary — it is an arbitrary (though measurable) point at which the ability to differentiate luminosity between two adjacent lines in a test image falls below an arbitrary standard. On either side of that there are still meaningful things going on in the captured image. So if you had a lens that produced (at some aperture and focal length) X lp/mm resolution, this does not mean that there would be no advantage seen in an image that contained areas in which there were subjects presenting more greater resolution challenges.
There are several other potential advantages, though again they mostly matter to people who make large, high quality prints:
1. The noise "grain" is smaller. In fact, in typical print sizes from a 50MP original, the individual pixels are so small in the print that noise is likely to be on a dimensional scale to small to see. In larger prints, say 24" x 36" and up, what noise is there will still be smaller grained.
2. There is a potential for smoother luminosity and color gradients, even in images that get pushed a bit in post by means of curves and so forth.
I don't mean to imply that a shooter who only shares jpg images online and in email or who prints no larger than 13" x 19" is going to see these things, but I don't think that the feature set of the 5DS is really targeted at that sort of user.
RustyBug wrote:
.. isn't it better to have the sensor outresolve the glass than the glass outresolve the sensor (i.e. moire', etc.).
Yup.
And the concept of "sensor out resolving the glass" (and vice versa) is a much trickier thing that many seem to acknowledge. Is it (or isn't it) important for the sensor to resolve at least the resolution provided by the best lens you use, measured at its best aperture, in the sharpest part of the image?
Peter Figen wrote:
Considering that more than one of their sample images are at f/11, those images are already hamstrung by diffraction. Not the best way to show off your new hi res sensor.
I think "hamstrung" might be overstating it a bit, but no one can argue that diffraction blur has no effect at f/11 on a full frame sensor body. You write that "more than one" of the images are at f/11 — but all one needs to do is look at the images that are not at f/11 to find out more.
Dan
|