Re: Bracketing DrT Style (Part II)

[P3D Michael Kersenbrock <michaelk@xxxxxxxxxxx>]

> >1/3  stop would yield better results.  
> 
> I have a feeling that 1/3 stop bracketing (two t's or one?) is a waste of
> film.  Who needs 1/6 fstop accuracy?

George,

You left off my "given".  I said that if one usually is off by at most
2/3rds of a stop, then using 1/3 stop bracketing would yield better
results.  In truth I should have said that 4/9-ths bracketing would
yield better results.  That is, a 2/9-ths stop worst-case error would
result with a 4/9ths bracketing and a 2/3rds stop maximum error from
center.  Makes the numbers messy when written that way though.  :-)

Given the assumption in my example, if you are off by the worst-case
2/3rds, or less, and you bracket +/- 4/9-ths, then the furthest one
can "miss" is by 2/9-th stop.  Given this same assumption, if one
brackets by a full stop, then one could miss by a full half
stop (instead of 2/9-ths).

In other words, if one usually guesses within 2/3rds stop, but wants it
better than that, the adding of two additional shots  yields 2/9-ths
stop accuracy!  One *COULD* use +/- one stop, and get 1/2 stop accuracy,
but why get 1/2 stop accuracy when 2/9-ths stop accuracy is the same
price with the same number of exposures?

> >As one develops a better "best guess" one's bracketting margins probably
> >get smaller.  
> 
> I'd rather see the frequency get smaller.  Is that what you meant?

I'm saying that if one brackets with ONLY a single triplet of exposures,
that the "spread" between the exposures shrinks as one gets better at
making "best guesses" (where the center exposure is).  Or should
shrink, for best results at the same cost.

I also suggested that the spread should be based both upon a person's 
personal history of "guessing" as well as the particular photographic 
situation.  The particular numbers used in the examples weren't 
suggested as typical.  If one thinks one guesses within 1 stop,
then bracketing by +/- 2/3rds stop would be optimal to yield a worst
case error of 1/3rd stop.


Mike K.

P.S. - Note that one may increase the spread for compositional purposes.  This
       is a different subject from bracketting to achieve a particular
       exposure optimally.   One may then bracket *each* of those compositional
       exposure choices.  :-)

P.P.S. - Depending upon the camera, one may also have entries in one's bracket formula
         that include things like the temperature or how long it's been since 
         the shutter clicked last time (especially for Kodak's with shutters that 
         haven't been cleaned since it was new :-).








------------------------------