Today's CCD or CMOS imagers are built (fabricated) on silicon wafer. It
means the imagers are 'flat'. Some of my understandings of these imagers
have limitations of the angle of the light entering the filter (RGB in case of
color) placed immediately above the sensor region. The maximum angle
from the perpendicular axis is said to be about 15 degrees (the angle
depends or the makers) for the light beam entering the filter. If the light
entering through the filter exceeds 10 degrees or so, the shading phenomenon
will occur. This phenomenon is represented such as with different color
image output between the center and at the edges of the area imager. The
shading phenomenon could be reduced by use of very high quality
aspherical lenses and ED lenses (probably much more complex than the
lens used in the standard E10 lens) or come up with a CPU or some sort of
arithmetic logic unit or digital signal processors (ALU or DSP) that can
calculate and predict the correct colors at the corners of the output images.
The application of convex shaped area imager including the filter, instead
of being flat imager, would solve the problem for the use of wide/super
wide angle lenses. But with today's technology, I think mass producing
the convex shaped imager and package is not practical or possible.
Jim Terazawa
Chip Stratton wrote:
> You'll notice that few digicams go wider than 35mm 'equivalent'. And you'll
> also notice that at that not so wide focal length, even the high end cameras
> have a fair amount of chromatic aberration.
>
> Apparently, since CCD sensors work best with light perpendicular to the
> chip, quality falls significantly as the angle of incidence becomes more
> acute, which is what happens with wide angle lenses as you move out from
> the center of the image. No doubt this problem is even more severe with the
> retrofocus designs used on wide angles for SLRs.
>
> I think we will be using film for the highest quality wide angle images for
> some time to come, at least until a different sensor chip technology comes
> along - or a better collimator design.
>
> Chip Stratton
> cstrat@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 >
|