Re: Nikon D80/200 - Canon 30d

On Jun 9, 9:34 am, Wolfgang Weisselberg <ozcvgt...@xxxxxxxxxxxxxx>
wrote:
new...@xxxxxxxxx <new...@xxxxxxxxx> wrote:
On Jun 8, 6:22 pm, Wolfgang Weisselberg <ozcvgt...@xxxxxxxxxxxxxx>
wrote:
http://www.clarkvision.com/imagedetail/evaluation-nikon-d200/index.html
That means:
ISO 100 200 400 800 1600
RN in DN 1.25 2.03 3.85 7.4 14.8
Ughh.
Serves me right for feeding the trolls.

I'd argue *I* am feeding a troll.


You'd lose. Again.



Whenever I claim something, you ignore or diss it.


Because you're totally and completely wrong. And I don't have nearly
as much spare time as you seem to, and I'd rather spend what little I
have shooting photos. OK?


When I show visual examples, the sources are out of their mind.

I pointed to several *major* electronics sites that state
unequivocally that CMOS sensors generate *more* noise than CCDs, not
less.

I also pointed to Ken Rockwell's site, which had pretty convincing
tests that demonstrated that differences in noise between the two
cameras is negligible. There is also an informative and 100% accurate
explanation of why measuring noise levels quantitatively is extremely
complicated, and that the software and equipment required to do it
properly do not even exist.

You on the other hand have pointed to Canon's marketing page and to
dpreview -- which basically just regurgitates Canon's marketing
hyperbole, uses rather dubious methodology, and grossly overstates the
differences in the images.

The only other "proof" you have is an obscure astronomy site, whose
results don't mean what you think they do.



When I ask you to compare apples to apples instead of apples to
oranges, it doesn't matter in "real photography" --- ignoring
my examples to the contrary.

Because it really doesn't matter. The fact that you're getting all
pathological over it makes me question your understanding of
photography. (You know, people used to have these same stupid
arguments about grain in film, which is also extremely difficult to
measure, and "THD" in stereo amplifiers).


When I show you the numbers --- hard, cold facts, derived from
physics and statistics --- you don't even try to understand
that they are not measurements, but claim that nobody does have
any idea how to do noise measurements and that the correct
instruments "don't tell you anything more than your eyes do"
(but do so consistently).

Hello? You just rejected the visual approach!

Yet your truths are simply based on you claiming them to be true.

No, they're based on reliable, readily-available information that can
easily be found by anyone who cares to look.


Photon noise is a fact (and I won't regurgitate physics and
statistics for you,

YOU are going to explain physics and statistics? That should be good
for a laugh.


You simply _assert_ them to be true, you don't even offer shreds
of proof. You do offer sources that disagree with each other,
you do show links that disclaim facts that you just asserted.


That's totally and comletely false. See above.


But while we're here, let's get real. You have absolutely NO IDEA
whatsoever what any of this really means, do you? Not a clue.


You completely missed the fact that that were not noise
measurements.


Umm, that's not what the page explaining the methodology says.

Basically, they take two images shot one right after the other, then
use ImagesPlus to "subtract" the two images so that all you are left
with is the noise from both images:

"Click the subtract button and the subtracted image is created. The
cross hair statistics tool will show the standard deviation, minimum
and other statistics. Be sure the minimum is above zero. If not,
increase the offset and click subtract again in the image math window.
The standard deviation represents the noise in TWO images. The noise
in one image is square root 2 less, so divide the standard deviation
by root 2 (1.4142) and record that value for both images."

See:

http://www.clarkvision.com/imagedetail/evaluation-1d2/


These are simple tests they are using to determine theoretical
limitations, and have little to do with what the human eye sees in
real-world photographs. In addition, one of the advantages of CMOS is
that you can apply NR circuitry onto the sensor itself, so that even
the RAW files have NR applied to them. In fact there's nothing to stop
the camera manufacturer from applying NR to RAW files from CCDs for
that matter. Which is one of the reasons why noise measurements are
largely meaningless.

.