At 2:25 AM +0000 9/20/03, olympus-digest wrote:
>Date: Fri, 19 Sep 2003 17:42:30 -0400
>From: Chuck Norcutt <chucknorcutt@xxxxxxxxxxx>
>Subject: Re: [OM] New E-1
>
>Julian said:
>IIRC, Oly said that the lenses were designed to support a sensor of 11 -
>12 Mp, which is the intended design limit of the system.
>- -----------------------------------------------------------
>
>Not sure if Oly has said it but Kodak certainly has. See:
><http://www.kodak.com/global/en/digital/ccd/sensorsMain.jhtml>
>and click on "New 4/3-type standard" in the upper right hand corner.
>
>That will take you to a PDF with a chart that clearly shows the
>resolution design limit for 4/3 lenses at 12 megapixels. While it
>doesn't say so, the curve extending from the current 5 megapixel sensor
>up to 12 megapixels I'm sure is intended to show the future development
>path for the 4/3 system.
>
>Curiously, this curve has some discreet points plotted on it at 3, 4, 5,
>6, 8 and 10 mp. After the 10 mp point the curve is dashed instead of
>solid. I can easily read design points into this that says the E-2 is 6
>mp, the E-3 8 mp and the E-4 is 10 mp. And like the OM line, there
>won't be an E-5 with 12 mp which would stretch the theoretical limits of
>the lenses.
>
>The chart also clearly shows that consumer grade DC's already have pixel
>sizes smaller than a 4/3 sensor with 12 megapixels. Therefore, the
>technical challenge isn't making small enough pixels but small enough
>pixels that also satisfy all the other system contraints.
There is a another issue to consider, dynamic range and noise floor. The
smaller the pixel the smaller the dynamic range, and that's what the figure is
trying to show. In brighter light, one can use smaller pixels to get the same
noise level.
The theoretical limit of the "Ideal Observer", achieved in scientific CCDs
anyway, is to be limited only by shot noise in either the photon flux falling
on the sensor, or in the electrons collected by that sensor. These both follow
Poisson statistics, which I described some time ago. Basically, the standard
deviation of the noise is equal to the square root of the average level. This
is inescapable, arising from the physics of light and of electron devices.
So, to achieve a given signal-to-noise level, one must collect a certain number
of photons (or electrons) per pixel. In bright light, a small pixel can do
this in a reasonable amount of time. In dimmer light, the pixel must be
larger, or one must accept a more noisy image.
Joe Gwinn
References:
The following two references expand upon the tradeoff between sensor resolution
(number of pixels for a given inage area) and dynamic range (larger pixels have
better dynamic range). Dynamic range is the illumination range from pixel
saturation to pixel noise floor, and is ultimately limited by the statistical
properties of light and the resulting photoelectrons (because they arrive in
discrete quanta, photons or electrons):
<http://www.kodak.com/global/plugins/acrobat/en/digital/ccd/papersArticles/ultimateSensor.pdf>
<http://www.kodak.com/global/plugins/acrobat/en/digital/ccd/papersArticles/interlineLargePixels.pdf>
This reference describes Kodak's brand new microlens CCDs, which are planned
for the 4/3 system. The point of the article is that they have achieved 85%
quantum efficiency, up from about 50%. Photographic film has a quantum
efficiency of 1% (slow film) to 10% (fast film).
<http://www.kodak.com/global/plugins/acrobat/en/digital/ccd/papersArticles/ulensAchievesPeakQE.pdf>
How to get from photometric units (visible light) to radiometric units (radiant
energy, not necessarily visible) is described in the following reference. CCD
sensitivity is specified in radiometric units, while cameras use photometric
units.
<http://www.kodak.com/global/plugins/acrobat/en/digital/ccd/applicationNotes/radiometryPhotometry.pdf>
The following reference tells how to measure the ISO sensitivity (=ASA) of a
CCD sensor, and concludes that the "base sensitivity" (explained in the ref) of
one of their 9-micron imagers is 100, and of a 6.8-micron unit is 180:
"Kodak's AF-16801CE, a 16 megapixel full-frame CCD image sensor with 9 micron
pixels, a Red-Green-Blue color filter array, and no microlenses has a base ISO
just over 100 assuming a demanding 170% reflectance tolerance for highlights.
This ISO is more than adequate for the studio lighting conditions under which
this sensor is typically used. Kodak's KAC-1310, a megapixel CMOS image sensor
with 6 micron square pixels, a Cyan-Yellow-Magenta color filter array, and
microlenses has a base ISO of 180 assuming 106% reflectance. This is more
appropriate for the lower light conditions often encountered by the snap shot
photographer."
<http://www.kodak.com/global/plugins/acrobat/en/digital/ccd/applicationNotes/ISOMeasurements.pdf>
< 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 >
|