[filmscanners] RE: Topic: RE: Newish Digital Tech

["Austin Franklin" <austin@darkroom.com>]

Hi Frank,

> this chip's [Foveon] potential
> implementation in film scanners (essentially a LINE-SCAN application)

Currently some scanners use tri-linear sensors, where each line is one of
the three colors, and others use either a tri-linear sensor or single
sensor, and vary the light.  There is nothing inherently about a Foveon line
sensor than what is currently used.

> The issue of the algorithms for organizing the Bayer pattern image into
> "the image" delivered by the camera, the interference patters, etc. --

What issue with the “algorithms”?  There are interference pattern issues,
but they are, for the most part, solved and therefore not an issue in
current designs.

> vs. the simplification and attendant improvement in speed

What improvement in speed?  You can shoot your DSLR in raw mode, where the
image data is not processed, or the processing is done in parallel with the
readout or with a secondary shot.  The processing time will be negligible,
if it isn’t already.

> And let us keep in
> mind that this chip technology is likely to evolve from its current
> nominal design/performance level, in the near term.

Very true.

> One of the interesting questions that comes to mind with the FOVEON and
> it 3 layered (G-B-R) arrangement -- is the relative "fill-factor"
> compared to a typical Bayer pattern RGB chip.  Fill factor is the
> proportion of the pixel area that is actually capturing light, with 100%
> being the upper limit -- and not achievable in practical
> implementations.

I am interested in what that number is for the Foveon, and aside from that,
how much light is blocked from going to the layers below by subsequent
layers.  The sensors must take up SOME of the light area...

> >From the perspective of my above observations, the FOVEON might not
> deliver any significant improvements to film scanners unless among its
> other attributes it has a superior fill-factor.

The linear sensors have a much better linear “fill factor”, such that it can
be nearly %100.  I do not see improving this in a scanner a significant
issue.

The advantage of using the Foveon in a scanner is it’s a one-shot, and would
make scanning go much faster.  Problem is the resolution and the
application. It may be good for scanning 4x6 prints, but not for negatives.

Regards,

Austin

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body