Re: [OM] E-10 Futuristic features?

Subject: Re: [OM] E-10 Futuristic features?
From: Paul Wallich <pw@xxxxxxxxx>
Date: Wed, 29 Nov 2000 00:10:24 -0500
At 9:53 PM +0100 11/28/00, David Irisarri <div2000@xxxxxxxx> wrote:

I´ve written to this list about E-10 and I´ve seen this
answer very interesting, so here it is.

I like you, have many OM lenses including the 180f2.0. An love it
dearly. A digital camera is however not a film camera and has many
different requirements. I can not answer for Nikon or others building
film digital bridge cameras, but this is the technology requirements
for a digital system. For example a digital/film lens SLR's using a
1 inch 2.7 megapixel CCD with a pixel pitch of about 12 microns.
(this varies according to whose spec you read from 11-13 microns.
The SLR lenses in general will resolve to 10 micron, on the surface
more than enough, but unlike film the CCD has valleys or wells, and
not a relativly smooth surface like film emultion.

This depends a lot on the CCD manufacturing method -- There are lots
of techniques to planarize a surface, and there are also techniques
(just like the ones used for putting on the filters) for putting microlenses
on the surface of the ccd to redirect light as needed.

The light ray angle of the lens will impact the CCDs ablility to deal
with the light, a 10-13 micorn pixel well can receive light from about
16-18 degrees. This means angles greater than the 35mm will not work.
(please not that the 20 is equivalent to a 35mm in all ways including
light ray angle in the area used by the CCD). When you go wider than
35mm your lens light ray angle increases, a 28mm (or equivalent) is
20 degrees or more. This is why we designed our E-10 lens the way it
is. When you get to higer density 4-6 million pixel CCDs the pixel
wells are smaller. A 1" 6 million pixel CCD is about 6 microns
reducing the angle of the light accepted to about 8 degrees, and a
2/3 inch 4 million pixel CCD will need light with less than 4-5
degrees of light ray angle. As we approch higher resolution CCD's to
allow for larger print sizes and more cropping the lens requirements
become ever more precise.

Not exactly precise -- rather, as you go to smaller pixels, if you don't
planarize, or if your wiring is too tall (yes, the wiring sits on top of
the sensor surface in most designs) you need a lens design with greater and
greater degrees of retrofocus. This would seem to argue, along with a lot of
other things, against using such tiny sensors.

There's also the supposed notion that the E-10 lens was designed for a
6-micron focus spot rather than the 10 microns of a typical film lens.
This seems suspicious to me. First, a 10-micron focus spot would pretty
much stop you from _ever_ getting better than 50 lines/mm resolution.
Second, I'm not clear on how a lens with a smaller aperture (because of
its much shorter focal length) can have better diffraction-limited performance.

Interchangability? The CCD/CMOS imager as I said is a rought surface
of ridges and wells. Dust and particulate matter falling on to this
suface can not be easily removed.resulting in dust specs on the
image. (check this out on some of the digicam chat rooms, it is a fact
of interchangable lens digicams). In a studio enviroment they wil work
ok if the camera is maintained in a clean
enviroment.                      The action of a focal plane shutter
(metal or cloth) will also
create micro dust particles that will attach them selfs to thecharged
surface of the CCD (resulting in dust specs). This is a challenge for
all high end digital manufacturers. One we as an industry will
eventually meet the right way, with the quality you have expessed for
the OM 4T system you have.

This seems to me to be pretty odd. I would think that a nice A/R coated
cover plate would take care of the problem for once and for all at minimal



Paul Wallich                                            pw@xxxxxxxxx

< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >

<Prev in Thread] Current Thread [Next in Thread>