[EMAIL PROTECTED] wrote:

> 
> Lemme see, I believe that that size of one CCD cell? in 1D is .0005 inch
> but as it is only sensitive to red green or blue youll need to look at
> values next to it also. If you take square of four, youll have enough
> information to get good colour estimate of the area and slightly better
> lightness info. Side of the square would then be 0.001 inch, which makes
> me wonder if you can produce much beyond that resolution anyway, even with
> spot on focus.
> 
> I hope I got the numbers right...
> 
>                                                         Karri

Your analysis seems reasonable, but, first, my calculation says that the
cell size is around .00043 inch which would lower the minimum resolution
your conclusions by itself.  Second, tests such as on dpreview.com
demonstrate that the 1D is capable of resolving a mininum of 1200 lines
per picture height, which turns out to be about .0006 inches *minimum
resolution (these tests indicate a horizontal LPH of 1400, and the
diagonal is higher than the test can detect).  As such, even in the
worst case, it should be capable of resolving detail below the .001/1.3
factor that would probably be the target for this "within the depth of
field" AF sensor calculation for acceptable AF accuracy.

Particularly when talking about resolution, whether in simple tests such
as dpreview's or in real world images, I think you should consider the
actual pixel count rather than one based on a 4x4 matrix of
pixels...except in the case of talking about pure red, blue or green
objects, in which case your point is pretty much valid, and it probably
does not detect a low enough frequency to hit the dof limit.  That is
why we see jagged edges on red and blue objects particularly...it does
much better on green though, since its resolution would be greater than
the red or blue items.  I have a nice photo of a cardinal that is hurt
significantly by this bayer pattern flaw.


Mike


-- 
Michael Shupe
M.J.Shupe Photography
Michigan Tech University
www.northernlightsgallery.com
*
****
*******
***********************************************************
*  For list instructions, including unsubscribe, see:
*    http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************

Reply via email to