Re: Sony A77 and A65 and Nex-7 Announcements Today
rico wrote:
I don\'t want to shake your closely-held beliefs, but this not a suitable subject for dogma. The math behind sampling is well established. I also see plenty of artifacts (color and monochrome) in the M9 example: trees, flagpoles, grass. The impact of a specific, moire-induced defect is not necessarily local, but can be global across the entire image, given the right input. The defect may be a false color or a false pattern of any type (random or periodic). Such defects cannot be erased with a Gaussian blur, image resizing, or resampling.
Moire can be global, sure, as I showed in the resizing example. Color artifacts can\'t. They are due to the way the CFA/demosaicing works and it\'s strictly local.
And of course such defects can be removed both by gaussian blur, resizing and resampling.
Look at the PP thread for many examples of methods of eliminating aliasing after the fact.
pdmphoto wrote:
The result of the 350D and M9 has as much to with the lenses used and post processing techniques, than the sensor. And, I do see all the color moire that has been mentioned. It even in the rock wall - which has patches of red, pink, and green in the grey (should all be grey). I shot with an SLR/c for over 5 years, so I know moire at 100% when I see it. It also very evident in the trees. It\'s not from downsizing for the web because its already at 100%.
The rock wall is granite so all those colors should be present. Here\'s a crop from a shot with a tele lens (100 MP on 5DII):
The claim that it\'s a question of differences in lens & PP when it comes to 350D vs M9 is about as absurd as saying that the difference between a bicycle and a ten ton truck is the tire pressure.
Aug 28, 2011 at 01:01 AM
Previous versions of denoir's message #9874635 « Sony A77 & A65 & Nex-7 Announcements Today »
