02-29-2012, 10:49 AM
Hmmmm... Wim, I realized that I saw this table before <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='
' />...
We will see many examples after D800 will be on the market. But still, the difference between MTF-50 and Rayleigh approach shows something that we can check over image examples. D800's new sensor is ~103 lp/mm (or let's say "D800E" to avoid AA filter interference). This means, diffraction effect on the resolution would be seen even @F/8 based on MTF-50 limits. OTOH if we apply Rayleigh, we see that @F/16 sensor outresolves the lens, and @F/11 lens resolution is clearly over the sensor's. The 100% viewed samples on the D800 technical guide, puts the limit in between, making things more complicated. Actually more or less the samples on page 13 justify the Rayleigh: "... you can see the grid lose definition as aperture is stopped down past f/11". But if we check Page 5, they show two crops @F/8 and @F/11, telling that there's a decrease on the system resolution (which can very hardly be seen). But F/11 -according to MTF-50- is 69 lp/mm... Quite below the sensor resolution... So the drop on resolution should be quite obvious @F/11. But it's not (at least not to me)...
I believe they should have to put an F/5.6 sample here... Because the decrease of system resolution is a relative thing considering the viewing critereias and even post processing activities. I mean one needs to know if it is a major drop from F/5.6 to F/8, and from F/8 to F/11 and F/16... Do we always view them 100%?... What about application of an USM to F/16 images?... So in the end, if our concern is the coping of resolution / contrast figures to real life examples and doing a verification, I guess we need to consider also other effects coming into play. But my personal opinion is that the Rayleigh approach seems to be more aligned with the practical usage.
Serkan
' />...We will see many examples after D800 will be on the market. But still, the difference between MTF-50 and Rayleigh approach shows something that we can check over image examples. D800's new sensor is ~103 lp/mm (or let's say "D800E" to avoid AA filter interference). This means, diffraction effect on the resolution would be seen even @F/8 based on MTF-50 limits. OTOH if we apply Rayleigh, we see that @F/16 sensor outresolves the lens, and @F/11 lens resolution is clearly over the sensor's. The 100% viewed samples on the D800 technical guide, puts the limit in between, making things more complicated. Actually more or less the samples on page 13 justify the Rayleigh: "... you can see the grid lose definition as aperture is stopped down past f/11". But if we check Page 5, they show two crops @F/8 and @F/11, telling that there's a decrease on the system resolution (which can very hardly be seen). But F/11 -according to MTF-50- is 69 lp/mm... Quite below the sensor resolution... So the drop on resolution should be quite obvious @F/11. But it's not (at least not to me)...
I believe they should have to put an F/5.6 sample here... Because the decrease of system resolution is a relative thing considering the viewing critereias and even post processing activities. I mean one needs to know if it is a major drop from F/5.6 to F/8, and from F/8 to F/11 and F/16... Do we always view them 100%?... What about application of an USM to F/16 images?... So in the end, if our concern is the coping of resolution / contrast figures to real life examples and doing a verification, I guess we need to consider also other effects coming into play. But my personal opinion is that the Rayleigh approach seems to be more aligned with the practical usage.
Serkan