>
> I really do not see why. Sure, the sensor should not reflect much light,
> but how hard can it be to compute the amount of light that has fallen
> onto the sensor when you know how reflective it is? Even if reflection
> is only .1% this seems verye feasable to me. For me It is still a big
> miss in all DSLR's, especially coming from Olympus.
>
> Someone care to elaborate?
>
It is actually quite difficult as the sensor is more "mirror like" than
film. Film has a very smooth scattering effect (matte surface) and is easy
to measure and gives consistent meter readings regardless of the position of
the light in the picture or whether the bright spots are small or large.
A sensor, being more mirror like, doesn't have this scattering effect and is
a high gloss surface. As such, the position and diameter of the bright spots
in the image are neither predictable or consistent.
For example go into your bathroom with a flashlight (torch). Turn the lights
off and look at the mirror. Now take the flashlight and point it at the
yourself in the reflection. Notice how bright the flashlight is? Now point
it elsewhere at the mirror. The reflected light is much dimmer because it
is no longer pointed at your face. Now, turn around and face the bathroom
wall. Do exactly the same thing by pointing the flashlight directly in front
of you and then to the side a bit. The brightness remains about the same
(depending on paint type).
This is what is happening inside the camera. A light-reading sensor in the
floor of the mirror-box is pointed at the imaging sensor. But since the
imaging sensor surface is mirror like, the light-reading sensor ends up
seeing the bottom of the mirror and none of the light coming through the
lens.
The only practical solution is to live-read the imaging sensor. I suspect
that the next generation of live-view imaging sensor will be shutterless.
Without a shutter, the sensor will receive a purge command at the start of
the exposure and then will dump to the interline rails at the end of the
exposure. Or the sensor will have a fixed read rate of 1000 frames per
second and exposures longer than that will be a cumulative stack of images
for the duration of the exposure. As the sensor requires no shutter and
light-levels are read dynamically off the main imaging-sensor, there is no
need for a parallel light-reading sensor system.
Or another possiblity is a hybrid sensor--one where there are pixels
scattered throughout the imaging-sensor and read and cleared at a high rate
as compared to the imaging pixels which are more traditional.
AG
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|