Archived posting to the Leica Users Group, 2013/11/07
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]Steve Barbour steve.barbour at gmail.com On Nov 7, 2013, at 9:11 AM, Paul Roark <roark.paul at gmail.com> wrote: >> ... You are by definition losing quality in those corners ... > > The a7r does have inward looking micro lenses. How they compare to the M9 > or M (which use different angles on them) will need tests to answer. On > the electronic side of the lens mount, however, I do not dismiss Sony. > > At this point, I don't see enough information in tests to tell whether the > Sony a7r (different than the plain a7) microlenses are better or worse than > Leica's. For now, I'd call that an open question. > > Assuming there microlenses are competitive, this raises the question for me > of whether in-camera correction profiles are necessarily better than doing > it in the computer in post processing. The corners are always going to be > darker than the center with any lens; vignetting to some degree is normal. > However, the corners of wide angle lenses can run into the dynamic range > limits of the sensor. That is, the corners get so much darker that they > are pushed down into unacceptably noisy values. Making a noisy image > lighter does not cure the problem. > > With a CCD, I don't see any theory that supports in-camera correction being > better than post-processing. Most of the noise if from the sensor site and > connection between the sensor and the processor. So for signal to noise > ratio, it's mostly game over by the time the data reaches the in-camera > processor. > > With a CMOS, there may be an advantage to in-camera processing. Where CMOS > gets its lower noise advantages with higher ISOs is by amplifying the > signal at the sensor site to reduce the impact of noise that is acquired in > the transmission of the data to the processor. However, for a CMOS > in-camera profile to have an advantage over post-processing, it would seem > that the camera would need to accomplish the vignetting correction via > increasing the effective ISO at the pixel site via differential > amplification. That is, the pixels would need to have increasing > amplifications as the pixels get further away from the center. I doubt > that the cameras are that sophisticated. I'm guessing all the corrections > are made in the camera's processor after the data is offloaded from the > sensor. If that is the case, then I doubt in-camera has any advantage over > post processing aside from convenience. > > Time and better testing is all that is going to answer our questions. The > concept of competition in digital bodies that will take our M glass, > however, is inevitable, and I expect it'll be wonderful news for the high > end, light weight market segment. Needless to say, the concept of mating > the best German optics to the best Japanese electronics intrigues me. me too Paul, and the rest was well stated (I think) :-) Steve > > Paul > www.PaulRoark.com > > _______________________________________________ > Leica Users Group. > See http://leica-users.org/mailman/listinfo/lug for more information