Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Jeff Medkeff
<medkeff@g...> wrote:
> schrochem wrote:
> 
> 
> > Now to pose a question I have been wondering about: if a sensor were 
> > made specifically for B&W, what would be changed?
> 
> I'd propose an answer to this question. Such a sensor would have no 
> Bayer mosaic, and would have a user-removable IR blocking filter (or no 
> blocking filter at all). In my ideal camera, it would have an 
> anti-aliasing filter (referred to as "blur" filter here earlier),
though.
> 
> This would send B&W photographers back to their color filters, of
course 
> - which I think would be a good thing in many respects.
> 
> Regarding expense: Such sensors are made by the thousands; Sony is an 
> especially prominent supplier. Such a camera would still be more 
> expensive than a color version, but I don't see any obvious reason for 
> it to be vastly more so.

The high cost is because of the low quantities. Look at the
Epson/Cosina/Voigtlander RD-1. Same sensor (or as near enough as makes
no nevermind) to the Nikon D70. Both are based on $200-300 film bodies
(Nikon N75, Bessa R3). But one costs $3000, the other $800. The D70
sells in the millions, so it's cheaper to manufacture. But more
importantly, it's cheaper to distribute, to design (better ratio of
R&D $ to sales $).

> Considering the sensor types that Nikon were 
> using about a year or year and a half ago, there were some off the
shelf 
> non-Bayer Sony sensors that looked very close to what was being used, 
> down to being pin-compatible. This suggests to me that Nikon might 
> pretty easily substitute one of these for the sensors they were using, 
> and issue a B&W version of an existing camera. Not sure if this could 
> easily be done by Canon or not.

Considering that Canon makes their own sensors. "hello, production.
Leave the Bayer filter off a thousand 20D sensors, we want to market
test a B&W camera".

> > If we knew what 
> > would be changed perhaps there is a filter set that can be used 
> > to "alter" what the sensor sees.....
> 
> For a "standard" filter to use with a Bayer mosaic, I would think that 
> you'd want a green-pass filter with an asymmetrical U-shaped extinction 
> curve - which would block, say, 75% of red and 50% of blue. And you'd 
> want the inverse of that filter; a green-block that passes red and blue 
> to those proportions.
> 
> Just scanning through one of my catalogs here I think this is going to 
> be pretty hard to find at reasonable cost.

Been there, done that.

Get the curves as csv tables (both Schott and Hoya will send them to
you that way) then write a relatively simple program to take your
desired curve and locate the best match from all the stacks of two
filters. That gives you tens of thousands of combinations, and one
will be a pretty good match.