on 2013-12-31 7:21 Ciprian Dorin Craciun wrote
     Looking at the two originals (linked below), even at 100% it's
harder to decide which is sharper.  However on closer inspection of
the previous two "helper" images I could conclude that 6226 has better
focus of the foreground, while 6229 has better focus of the distant
shore.

       
http://data.volution.ro/ciprian/bdf753de58c560bc/6d20367506baa6fb/_igp6226/preview-rgb.ppm.jpg
       
http://data.volution.ro/ciprian/bdf753de58c560bc/6d20367506baa6fb/_igp6229/preview-rgb.ppm.jpg

maybe this isn't a great example — while i agree it's hard to tell visually which has better focus on the rocks (and thus one could ask if it really matters), i can easily see the difference in focus on the trees; thus, given your intent, 6229 is the obvious choice

    For other three tough decisions see the following "siblings"
bracketed in a hurry at 0.7 EV: [...]

    Their exposure (seems the meter was dead-on):  [...]

this is also pretty obvious visually, and i'd expect a histogram to make it more obvious; what i can't tell is how much exposure latitude you have in your sources — it may be that any of the three could be adequately corrected (what camera are you using?)

    Their sharpness (seems the second one is less fuzzy, however not by much):  
[...]

i can see this visually, and normalizing the exposures would probably make it easier to see; i'm not convinced of the usefulness of your focus tool, especially since you said your stock tools already have some sort of "in focus highlight"

one could also ask — how often do you print large? how often do you crop small bits of an image? these are the cases where top sharpness matters; otherwise no one but you will see the difference; and i second Attila's recommendation to study how to sharpen for your intended output — mastery of sharpening can make a bigger difference in the final result than the tiny differences you have trouble seeing at 100%


--
PDML Pentax-Discuss Mail List
[email protected]
http://pdml.net/mailman/listinfo/pdml_pdml.net
to UNSUBSCRIBE from the PDML, please visit the link directly above and follow 
the directions.

Reply via email to