Also, I meant to point out earlier that I didn't consider "painterly" to
be a valid part of the judging criteria unless I misunderstand the word.
Moose was specifically looking for an image that best displayed a
notion of detail in the leaves. My notion of "painterly" is what we see
on an artists canvas which, while it may be beautiful, never represents
the detail one sees even in a not very sharp photograph.
I think, given the limited judging criteria, that one can rank the
images with N-1 comparisons.
Chuck Norcutt
usher99@xxxxxxx wrote:
> I like the simplicity/efficiency of that approach but I am not sure it
> will properly rank the images or pick a
> valid winner unless there is a UCLA dynasty caliber image among the
> choices. Sudden death elimination may be too harsh for a carefully
> crafted image alternative. As pointed out the comparrison image may
> influence the choice.
> Sooo, a spunky underdog Gonzaga image may be eliminated in an early
> round due to an unfortunate pairing early on.
> I would favor the pair-wise comparison with a multiplicative factor of
> 1 to 3 , with 3 representing a marked difference in favor of the
> winning
> image in the pair--then add up the points. I think with another
> 10-15min of study I could determine a final four but as you correctly
> point out,though I could ascertain nuanced differences among the
> images, proper articulation of that is beyond my abilities.
> Getting too close to the monitor can make processing artifacts more
> apparent and is likely not fair. I think viewing from a standard
> distance should help.
> Soo, let the Moose March Madness begin.
>
> Mike
>
>
>
>
>
>
>
>
>
> I didn't see the situation as such a big combinatorial problem if you
> could (as I mentioned earlier) pick a base image and a comparison image
> and have them overlaid as Moose does sometimes to simplify comparison
> of
> the pair. Choose image #1 as the base and #2 as the comparison. If
> the
> base is better leave the base at #1 and move the comparison image to
> #3.
> If the comparison image is better move the base to #2 and then the
> comparison to #3. This way there are only N-1 paired comparisons or 10
> vs. 55. This assumes, of course, that you do not change your judging
> criteria mid-way through the process (eg; "edgiest" vs "most
> painterly") :-)
>
> Chuck Norcutt
>
>
> usher99@xxxxxxx wrote:
>> Wow,
>> Finished up relatively early at work, but this is a tough
>> determination. Difficult form a processing vantagepoint and for
>> experimental design in picking a favorite. Should be 55 pair-wise
>> comparisons (11 items taken 2 at a time) The adjacent image seems to
>> influence the impression as well. 2 seems the the brightest and
>> "edgiest"--perhaps more LCE + sharpening? 4 and 9 seem more
> painterly
>> with 4 clearly different than all the others. The latter two have
> less
>> texture (?microcontrast) but not unpleasant. I wouldn't pick them as
>> my favorite unless I am right on top of the screen.
>>
>> I can't keep the nuances straight on all the choices w/o much study
>> at this hour--so can't pick a clear winner.
>>
>> MMM (mostly manual focus, Mike)
>>
>>
>> Today I decided to try some different approaches to the problem.
> Anyone
>>> with sharp eyes is invited to peruse a bunch of alternatives and
> give
>> me
>>> their opinions. Those who can't MF focus anymore may see little
>>> difference between some (most?) of them.
>>> <http://www.moosemystic.net/Gallery/tech/Foliage/DownTexture.htm>
>>>
>>> Any help appreciated.
>>>
> --
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|