CCD chips VS CMOS chips

frederick wrote:

The samples that are now widely distributed and were slammed
by doubters as "fakes" because of incomplete exif are also
available as original jpegs with intact exif.

In other words, soneone did a bad job faking the EXIF values from
a 40D and someone else did clean up after him? :-)

I'll rather wait till independend 3rd party shots and evaluations
are in before I form an opinion in either way.

-Wolfgang

David J Taylor wrote:

Well, Wolfgang, I would hope that the NR removed the colour /noise/, but
not the colour itself! G I agree that for a fair comparison, any NR
processing should be the same, but I am also happy to compare different NR
methods. As the eye is less sensitive to some aspects of colour (e.g.
spatial resolution), more colour NR may lead to a picture which is
subjectively better. Some small-sensor cameras already do this.


Given the choice, I would rather remove noise in software rather than in
camera .... so I hope that NR is an OPTION. It seems to be an option on Nikon
cameras but I am not sure about it on Canon. Ever more advanced methods of
noise reduction become available should not obsolete your camera, just your
software. Having said that, my Nikon D200 should last me until at least the
release of the D400 ... and likely later.

--
Thomas T. Veldhouse

We have more to fear from the bungling of the incompetent than from the
machinations of the wicked.

wrote:
Check out these two links:

http://www.imaging-resource.com/PROD...30DINI3200.HTM
http://www.imaging-resource.com/PROD...200INI3200.HTM

The Nikon (CCD) image has a bit more noise, but it also captures more
detail (evident in the model's face and hair, and the ficus tree in
the background). Also, the Canon image is overexposed slightly, which
tends to reduce noise somewhat. But there really isn't much difference
between the two.

Well, the main problem is that these are 2 different
situations (the model has moved, the light on the skirt is
different ... so these are somewhat hard to compare.

I'd say that the D200 does supress colour in the dark parts
of the bouquet, giving gray/black areas, whereas the 30D does
render even the darkish leaves there as green. But again,
since the position of the bouquet and light is not identical,
it's somewhat hard to compare.

See e.g. the leaf directly over the base of the thumb in the
200D image, half rolled into itself, and compare to the
equally dark leaf upside down above the base thumb in the 30D
image, the 2 leaves right of it or the the folded leaf above the wrist.

Just find a camera that fits your budget, handles well and takes the
photos you want. Don't lose too much sleep over the type of censor.
They're both good.

Yep. Unless you need the very last bit of light, in which case
you need a lot more money ...
http://www.fairchildimaging.com/prod...regrine486.htm
.... start at $16k (for a front-illuminated scientific grade 3 one)
and end at $95k (back-illuminated quality grade 1). Oh, and at
61mmx61mm you need larger (medium format) lenses. And a camera
body to attach them to, and anything like AF you might want. And,
yes, cooling the sensor to -50°C gives interesting things like,
say condensing water and electronics to have headaches about.

-Wolfgang

Wolfgang Weisselberg wrote:
wrote:
Check out these two links:

http://www.imaging-resource.com/PROD...30DINI3200.HTM
http://www.imaging-resource.com/PROD...200INI3200.HTM

The Nikon (CCD) image has a bit more noise, but it also captures more
detail (evident in the model's face and hair, and the ficus tree in
the background). Also, the Canon image is overexposed slightly, which
tends to reduce noise somewhat. But there really isn't much
difference between the two.

Well, the main problem is that these are 2 different
situations (the model has moved, the light on the skirt is
different ... so these are somewhat hard to compare.

I'd say that the D200 does supress colour in the dark parts
of the bouquet, giving gray/black areas, whereas the 30D does
render even the darkish leaves there as green. But again,
since the position of the bouquet and light is not identical,
it's somewhat hard to compare.

See e.g. the leaf directly over the base of the thumb in the
200D image, half rolled into itself, and compare to the
equally dark leaf upside down above the base thumb in the 30D
image, the 2 leaves right of it or the the folded leaf above the
wrist.
[]
-Wolfgang

It's very difficult to say, with the different focussing, different
exposure, and different (to me) contrast in the two images. It seems to
me thtat the darkers areas on the Nikon are being pushed off the lower end
of the dynamic range, rather than simply loosing their saturation. Pity
there wasn't more control when taking, and a testcard in the scene!

Cheers,
David

On Aug 27, 10:11 am, Wolfgang Weisselberg
wrote:
David J Taylor wrote:

Wolfgang Weisselberg wrote:
Canon is currently the "low light" leader, they did wonders
with CMOS.
That may be "was the leader" - the new Nikon cameras may have reversed the
leadership. Awaiting 3rd party tests.

I'd be surprised, but happy enough.
A lot depends on the testing --- a NR that removes colour
in some circumstances might _look_ cleaner, but ...


I would not be surprised if the D3, at least, had very low noise.

Wolfgang Weisselberg wrote:
frederick wrote:

The samples that are now widely distributed and were slammed
by doubters as "fakes" because of incomplete exif are also
available as original jpegs with intact exif.

In other words, soneone did a bad job faking the EXIF values from
a 40D and someone else did clean up after him? :-)

For sure it would have been a bad idea to use the 40d to
fake d3 images. There are the wrong size, they are much
noisier (similar to 400d) and freely available from sites
such as DPReview.


I'll rather wait till independend 3rd party shots and evaluations
are in before I form an opinion in either way.

Sure, an added problem is that in some of the D3 test shots,
a long telephoto at f2.8/f4 was used, DOF is so shallow that
there's not so much in pin-sharp focus.
I don't really have an opinion, but an expectation based on
what I've seen, and what others have said, that Canon now
takes No2 position behind Nikon on quality of it's high-end
cameras, as well as it's lower priced dslrs.

cjcampbell wrote in
ups.com:

I would not be surprised if the D3, at least, had very low noise.

If it starts at ISO 200 (100 is a special extension, perhaps because of
missing highlight headroom), then that could because of a higher quantum
efficiency, which, combined with the full frame, could collect a lot of
photons, for low shot noise. That doesn't tell us much about read noise,
however.



--


John P Sheehy

wrote in news:1188349251.823439.141220
@r29g2000hsg.googlegroups.com:

The Nikon (CCD) image has a bit more noise, but it also captures more
detail (evident in the model's face and hair, and the ficus tree in
the background). Also, the Canon image is overexposed slightly, which
tends to reduce noise somewhat. But there really isn't much difference
between the two.

Did you fail to note that the foliage is what is in focus in the Canon
image, and that the face is what is in focus in the Nikon image?

Did you also fail to notice a lot less noise/detail reduction in the Canon?

The in-focus areas of the Canon image are superior to the in-focus areas of
the Nikon image.

The Nikon image breaks up much faster as you get into the very darkest
areas.

--


John P Sheehy

"David J Taylor"
wrote in
k:

It's very difficult to say, with the different focussing, different
exposure, and different (to me) contrast in the two images. It seems
to me thtat the darkers areas on the Nikon are being pushed off the
lower end of the dynamic range, rather than simply loosing their
saturation. Pity there wasn't more control when taking, and a
testcard in the scene!

The D200 has about 52 ADU read noise at ISO 3200; the 30D has less than 10
ADU. The D200 has about 5.5x as much read noise.

You can see it and its reduction all over the darker parts of the D200
image.

--


John P Sheehy

John Sheehy wrote:

wrote

The Nikon (CCD) image has a bit more noise, but it also captures more
detail (evident in the model's face and hair, and the ficus tree in
the background). Also, the Canon image is overexposed slightly, which
tends to reduce noise somewhat. But there really isn't much difference
between the two.


Did you fail to note that the foliage is what is in focus in the Canon
image, and that the face is what is in focus in the Nikon image?

Did you also fail to notice a lot less noise/detail reduction in the Canon?

The in-focus areas of the Canon image are superior to the in-focus areas of
the Nikon image.

The Nikon image breaks up much faster as you get into the very darkest
areas.

The bouquet looks much better on the Canon, the face & ficus look much
better on the Nikon. The Canon shot is considerably more brightly
exposed. Not really a very useful comparison in the end I think. 3200
would probably be better done underexposing & post-processing anyways.
1600 would be more meaningful.

--
Paul Furman Photography
http://edgehill.net
Bay Natives Nursery
http://www.baynatives.com