05-23-2012, 02:34 PM
Klaus, I appreciate your responses, thank you.
However, you still haven't explained where you believe the connection to "RAW profiles" is. I think there is none and it may be good to revise that statement.
[quote name='Klaus' timestamp='1337667093' post='18420']
Just think of the AA filter as as opal/milk glass.
[/quote]
It is better to think of it as two layers of bifringent material. <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='
' />
[quote name='Klaus' timestamp='1337667093' post='18420']
And, frankly, I'm just reading the measured results here - I've not created a new hypotheses that I need to proof.
[/quote]
True. However, you are publishing measurement results and should therefore have an interest in them being as accurate as possible (within reasonable boundaries).
Your MTF results are given in terms of "lp/mm" and strictly speaking the respective numbers should hence be independent from the sensor used, as long as the lens does not outresolve the sensor. Most lenses will easily outresolve most sensors in the centre at their optimal aperture, so we should be expecting better performance as sensors with higher pixel pitch become available. However, when the lens does not outresolve the sensor (borders, high f-ratios) it should yield the same results independently of the sensor's pixel-pitch.
I appreciate that the latter may not be as easy to achieve as it seems (in particular when sensors like that from the K10D are involved) but when a retest yields results that appears to be in conflict with earlier tests then I believe it is worth investigating potential causes in order to eliminate any systematic problems with the testing procedure. Thus I believe you should be trying to create theories and attempt to support them with evidence; in the interest of reliable results.
However, you still haven't explained where you believe the connection to "RAW profiles" is. I think there is none and it may be good to revise that statement.
[quote name='Klaus' timestamp='1337667093' post='18420']
Just think of the AA filter as as opal/milk glass.
[/quote]
It is better to think of it as two layers of bifringent material. <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='
' />[quote name='Klaus' timestamp='1337667093' post='18420']
And, frankly, I'm just reading the measured results here - I've not created a new hypotheses that I need to proof.
[/quote]
True. However, you are publishing measurement results and should therefore have an interest in them being as accurate as possible (within reasonable boundaries).
Your MTF results are given in terms of "lp/mm" and strictly speaking the respective numbers should hence be independent from the sensor used, as long as the lens does not outresolve the sensor. Most lenses will easily outresolve most sensors in the centre at their optimal aperture, so we should be expecting better performance as sensors with higher pixel pitch become available. However, when the lens does not outresolve the sensor (borders, high f-ratios) it should yield the same results independently of the sensor's pixel-pitch.
I appreciate that the latter may not be as easy to achieve as it seems (in particular when sensors like that from the K10D are involved) but when a retest yields results that appears to be in conflict with earlier tests then I believe it is worth investigating potential causes in order to eliminate any systematic problems with the testing procedure. Thus I believe you should be trying to create theories and attempt to support them with evidence; in the interest of reliable results.