Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "John Moody"
<moodymz3@y...> wrote:
>
> I don't see any use for it by fine art photographers.  The optical
path is
> messy; adds losses and distortions by splitting the incoming rays
into many
> more rays.

Actually, the "losses" are tractable, the information is recovered in
the recombination. And it reduces distortion, virtually eliminating
sensor angular effects like purple fringing, and adding ways of
compensationg for vector seperable distortions such as comma or
spherical aberration.

> A firmware update, to add focus bracketing on existing dSLRs
> would be of more use to us, IMO, and cost nothing.

And it would be worth every penny of... nothing.

DSLRs with "screwdriver" drives (Nikon, Pentax, Minolta) can't focus
bracket at all because of mechanical backlash and the low precision of
the distance reporting from the lenses. Micromotor USM (most low to
mid cost Canon lenses, and all Oly 4/3 lenses) have the same
restriction. Ring USM (Higher end Canon, some Nikon AF-S) can do it,
but slip sometimes screws it up, you can't count on it.

And yes, I'm "that" Joe Wisniewski. I have done this before for a four
camera array scanning telescope. We ended up having to do a "return to
zero" between each successive shot. Slow, and it generated a lot of
torque vibration with the whole array focusing.

As far as using multiple "focus stackin" in day to day photography,
it's not all that easy (I speak from years of experience). You can't
use it on living subjects (obviously) because they won't align in the
different layers. Even breathing throws the stack off. You can't use
it on a landscape unless the wind is very calm that day (forget it if
there's a lake or pond in the scene).

And you can't really build it into a DSLR, because there's so much
math. And it's not "pretty" math. The math to combine focus layers is
"soft", there are decision algorithms, in addition to a lot of "hard"
math dealign with the alignment (because at the pixel level, camera
and lens shift between shots, even on a sturdy tripod) and the
resizing (because refocusing a lens changes the image coverage angle,
and may even cause parallax errors due to movement of the entrance pupil).

Plenoptics is all "hard" math, no decisions, just calculation.

> On the surface it looks very interesting, but after looking at the
actual
> method used, I doubt it will have any significant commercial success.

Actually, from a technology forecasting standpoint, it looks like the
FOM is holding, and plenoptics will be commercially viable within 10
years.