Jeff Keller wrote:
>Perhaps one of the reasons you chose a C*non over the E-1 ? ;-) Actually
>the E-1 really does have the pixels to support 11MB raw files.
>
>
I don't know what that means. Are some camera's pixels somehow bigger or
meatier than others?
Both the E-1 and the 300D use Bayer pattern sensors, with 25% reading
blue, 25% red and 50% green.
<http://www.dpreview.com/learn/?/key=color+filter+array> Both sensors
are also 12 bit output, so the RAW data is 12 bits per pixel
<http://www.dpreview.com/learn/?/Glossary/Digital_Imaging/RAW_01.htm>.
Thus, the maximum actual data from an E-1 sensor is 2560x1920 =
4,915,200 x 12 bits / 8 bits per byte = 7,372,800 bytes. So what is in
the extra 3 mb of an E-1 RAW file? One answer may be the inclusion of
lens shading/vignetting info from the lens at the pixel level? I seem to
recall that lens shading correction, which is prohibitively slow in
camera, can be applied later in RAW processing. 4 bits of luminance
adjustment data per pixel would account for the difference. Why they
would do that, rather than just include the lens data for application at
the pixel level in processing, I don't know. Just speculation.
For the 300D, 3072x2048 = 6,291,456 x 12 / 8 = 9,437,184 bytes. Not only
that, but 300D RAW files include an imbedded JPEG. So where do the extra
bytes disappear to? The asnwer is almost certainly lossless compression
of the RAW data.
Now I'm not sure how much practical difference there is between a 4.9
and a 6.3 mp sensor; not much, I believe. However, in the rather small
matter of RAW output storage efficiency, the E-1 uses more storage for
smaller images than does the 300D. That's all I was saying.
Moose
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|