I would suspect that lowering heat with more efficient electronics would
be one way to decrease noise. One of the problems with the early Olympus
digital cameras, and maybe why Canon, with CMOS sensor is able
to do better, since CMOS circuits tend to be lower power, and as Chuck
says, lower noise electronics in the backend system helps. I think the silicon
noise sources dominate more over quantum efficiency, but I'm just
guessing. CMOS analog sampled data circuits can do a lot of signal
processing before being digitized that cannot be done easily in other
silicon technologies, on the sensor chip. CCD's have to pump the charge
out, while a CMOS pixel can have dedicated electronics per pixel. CCD's
should be lower noise than CMOS, but the difference in the peripheral
electronics probably makes the difference.
All things being equal, though, the larger the sensor the lower the
noise. Same as film with a given grain noise. I still think Olympus has
boxed themselves in with the 4/3 system. So much so, I have now
switched to the Borg 5D. Olympus still gets some things better though
in terms of camera control and user interface. But just like no auto-focus
technology in the OM days, no IS or VR today. Maybe with the 4/3 sensor
they can do something like Konica-Minolta. A smaller sensor would help
with that technology, and it would extend the really nice Olympus lenses.
Wayne
At 11:04 AM 10/24/2005, you wrote:
>ScottGee1 wrote:
>
>>
>> Following that line of logic, packing more pixel sites onto the same
>> size sensor *should* result in lower quality and more noise even
>> though the resolution is increased. But apparently that ain't true.
>>
>> I gotta lot to learn . . .
>-----------------------------------
>
>That is true in some cases and would be true in all cases except that
>technology marches on. A physically smaller pixel has less area with
>which to capture photons and the fewer the photons the lesser the ratio
>between signal (photons) and noise. Noise is generated within the
>sensor's electronic circuits <snip>.
>
>I'm not sure if any sensor manufacturer has actually figured out how to
>increase the quantum efficiency of a pixel but they have certainly made
>advances in producing quieter circuitry and employing various tricks to
>differentiate signal and noise and subtract the noise. <snip>
>
>Chuck Norcutt
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|