Usually it's some simple algorithm like bicubic, due to its speed and
efficiency. More advanced computer-based editor programs use one of the many
Lanczos resampling algorithms which would retain more detail, and hence more
noise. That comes at the price of processing time, something which a camera
tends to avoid. In this case, a camera's processors would be incapable of
executing a Lanczos-8 resampling routine in the time needed. It even taxes good
computers at times.
See this thread at dpreview which might answer most of your questions:
http://forums.dpreview.com/forums/readflat.asp?forum=1010&thread=29187421&page=1
That discussion covers most of what you ask and wonder about.
The pixels should not actually increase in size. The sensor itself is
just using fewer pixels to capture the image so the size of pixel
should be the same.
On some cameras (particularly high end video cameras used on
microscopes) you can "bin" pixels to increase sensitivity - ie you can
use two pixels on the sensor to create one pixel in the image (or 4 to
1) etc....The resolution decreases but the sensitivity increases. I
would think however on your Ricoh, it is simply leaving out the
information from the non-used pixels though.
>The pixels should not actually increase in size. The sensor itself is
>just using fewer pixels to capture the image so the size of pixel
>should be the same.
No. A digital camera only uses fewer sensor pixels when sending a live-view feed
to an EVF or LCD on the camera, or sometimes when using the larger sensor for
capturing lower-resolution video-frames (often this is done with pixel-binning
instead though, for better low-light low-noise performance). All cameras that
allow you to choose a smaller image size for the resulting stilll-frame JPG file
does so by using the full sensor data and then downsampling it.
Depending on the downsampling algorithm used, this can blur fine details (e.g.
bicubic) and also blur/average out some noise too in the process. If you have no
need for higher resolution images, and you need to shoot at higher ISOs that can
be noisier, you can do some extra in-camera noise-reduction by downsampling,
choosing a smaller image size. It's not the best way to do it but for some
people's needs it might be all they care for. That's perfectly fine to do so.
As has been proven time and time again, resolution, quality, noise, none of that
really matters in an award-winning image. The content of an image will win out
over those things each and every time. I once had a 1024x768 image from an early
digital camera. The subject so striking that it even held up to being printed at
13"x18", after some well-executed upsampling to remove the garishly huge pixel
blocks. The image having so much impact that nobody cared about the ultra-soft
edges (which were part of the original to begin with to some extent). They never
even noticed the softness in the photo because the subject was more important to
their eye and mind. The lower quality hidden behind the impact of the subject.
There are no absolutes when it comes to photography. The only possible absolute
in photography is that the image will hold its own at any resolution, or it
won't at all at any resolution.
:-)
It's heresy only to a bunch of online techno-geeks; devoid of any contact with
reality; who can't, won't, and don't fully comprehend this wonderful quote:
"We often think that when we have completed our study of 'one' we know all about
'two', because 'two' is 'one and one'. We forget that we have still to make a
study of 'and'." - A. S. Eddington
No, I agree with Vern on this point. Of course that's no reason to use a
lesser camera if they want to take better photos.
--
Paul Furman
www.edgehill.net
www.baynatives.com
all google groups messages filtered due to spam