01-21-2012, 12:22 PM
[quote name='Frank' timestamp='1327115968' post='15078']
Hi Wim, thank you for your answer. I understand that the tolerance for tilt should be very small, especially when the lens is focused at infinity. But how about the tolerance for the shift? If the optical axis of the lens is off set from the center of the sensor but it is still orthogonal to the sensor plane, then I would expect that at the four corner regions of the sensor the decrease in resolution and the vignetting would be nonsymmetric. How about the tolerance in this case? I expect that the tolerance in this case would be larger than the tolerance in tilt, but I would like to check with you. Does a lens factory control the tolerance in shift as stringently as in tilt?
Regards,
Frank
[/quote]
I would expect that shift is simply left within design tolerances. I.e. I mean a lens is circular, so everything is theoretically centered. Likewise, lens mount and sensor are mounted in the place where they are calculated to be. So, I don't think much is adjusted here. Based on hard vignetting testing, indeed corners are not exactly the same for different lenses. http://photonius.wikispaces.com/Vignetting
Hi Wim, thank you for your answer. I understand that the tolerance for tilt should be very small, especially when the lens is focused at infinity. But how about the tolerance for the shift? If the optical axis of the lens is off set from the center of the sensor but it is still orthogonal to the sensor plane, then I would expect that at the four corner regions of the sensor the decrease in resolution and the vignetting would be nonsymmetric. How about the tolerance in this case? I expect that the tolerance in this case would be larger than the tolerance in tilt, but I would like to check with you. Does a lens factory control the tolerance in shift as stringently as in tilt?
Regards,
Frank
[/quote]
I would expect that shift is simply left within design tolerances. I.e. I mean a lens is circular, so everything is theoretically centered. Likewise, lens mount and sensor are mounted in the place where they are calculated to be. So, I don't think much is adjusted here. Based on hard vignetting testing, indeed corners are not exactly the same for different lenses. http://photonius.wikispaces.com/Vignetting