EOS 20D Low Noise

Alfred Molon wrote:

des wrote:

According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..


How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?

o New CMOS imager
o New processing firmware.

How can we think improvements have topped out with the 10D/300D?

Phil

Kibo informs me that Alfred Molon stated
that:

des wrote:
According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?

Better processing on the analog side, presumably, as with the 1DMkII.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

Kibo informs me that Alfred Molon stated
that:

des wrote:
According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?

Better processing on the analog side, presumably, as with the 1DMkII.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

Easy - two years newer technology, which means larger micro-lenses, and also
less transistors in the pixels. The first means less light is lost between
the pixels, the second means the physical size of the pixels can be larger,
despite the pixel pitch being closer together.

Easy - two years newer technology, which means larger micro-lenses, and also
less transistors in the pixels. The first means less light is lost between
the pixels, the second means the physical size of the pixels can be larger,
despite the pixel pitch being closer together.

wrote:
Kibo informs me that Alfred Molon stated
that:


des wrote:

According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?


Better processing on the analog side, presumably, as with the 1DMkII.

I'm doubtful. If it were true I'd be really tempted. The 1D-II has a
much larger sensor and the same imaging chip as the 20D. At ISO 1600
is it as good as the 10D at 400?

http://www.dpreview.com/reviews/cano...kii/page18.asp
http://www.dpreview.com/reviews/canoneos10d/page18.asp

The different size and layout of the crops makes it a bit hard to judge
from these pages but I'd say not. I think the 1D-II at 1600 might be
fairly close to the 10D at 800, but not 400.


Rob.
--

wrote:
Kibo informs me that Alfred Molon stated
that:


des wrote:

According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?


Better processing on the analog side, presumably, as with the 1DMkII.

I'm doubtful. If it were true I'd be really tempted. The 1D-II has a
much larger sensor and the same imaging chip as the 20D. At ISO 1600
is it as good as the 10D at 400?

http://www.dpreview.com/reviews/cano...kii/page18.asp
http://www.dpreview.com/reviews/canoneos10d/page18.asp

The different size and layout of the crops makes it a bit hard to judge
from these pages but I'd say not. I think the 1D-II at 1600 might be
fairly close to the 10D at 800, but not 400.


Rob.
--

Kibo informs me that Rob Davison stated
that:

wrote:
Kibo informs me that Alfred Molon stated
that:


des wrote:

According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?


Better processing on the analog side, presumably, as with the 1DMkII.

I'm doubtful. If it were true I'd be really tempted. The 1D-II has a
much larger sensor and the same imaging chip as the 20D.

No, it's a different sensor. The 20D is a 1.6x crop, vs the 1DMkII 1.3x
crop. The on-sensor processing system is different as well.

At ISO 1600
is it as good as the 10D at 400?

According to Canon & Rob Galbraith, yes:
http://www.robgalbraith.com/bins/multi_page.asp?cid=7-6458-7153-7154
"Canon is promising that, because of the newly-developed noise
processing and other sensor design changes that 20D noise levels at ISO
1600 will roughly match the 10D at ISO 400"
If that's true, I'm very impressed, because I use my 10D a *lot* at ISO
400-1600, & ISO 1600 on the 10D is quite usable for my photography, &
ISO 400 is very good indeed.
The single 'gotcha' I can find on the 20D so far is that it's buffer in
RAW mode is about the same as on the 10D, which I suspect is an
artifical limitation to prevent it from competing with the 1DMkII.

http://www.dpreview.com/reviews/cano...kii/page18.asp
http://www.dpreview.com/reviews/canoneos10d/page18.asp

The different size and layout of the crops makes it a bit hard to judge
from these pages but I'd say not. I think the 1D-II at 1600 might be
fairly close to the 10D at 800, but not 400.

Bear in mind that DPreview used their newer testing method on the
1DMkII, so you can't compare the 10D & 1DMkII results directly. If you
look at the top of the first link you quoted, you'll find that they have
a notice there pointing that out.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

Kibo informs me that Rob Davison stated
that:

wrote:
Kibo informs me that Alfred Molon stated
that:


des wrote:

According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?


Better processing on the analog side, presumably, as with the 1DMkII.

I'm doubtful. If it were true I'd be really tempted. The 1D-II has a
much larger sensor and the same imaging chip as the 20D.

No, it's a different sensor. The 20D is a 1.6x crop, vs the 1DMkII 1.3x
crop. The on-sensor processing system is different as well.

At ISO 1600
is it as good as the 10D at 400?

According to Canon & Rob Galbraith, yes:
http://www.robgalbraith.com/bins/multi_page.asp?cid=7-6458-7153-7154
"Canon is promising that, because of the newly-developed noise
processing and other sensor design changes that 20D noise levels at ISO
1600 will roughly match the 10D at ISO 400"
If that's true, I'm very impressed, because I use my 10D a *lot* at ISO
400-1600, & ISO 1600 on the 10D is quite usable for my photography, &
ISO 400 is very good indeed.
The single 'gotcha' I can find on the 20D so far is that it's buffer in
RAW mode is about the same as on the 10D, which I suspect is an
artifical limitation to prevent it from competing with the 1DMkII.

http://www.dpreview.com/reviews/cano...kii/page18.asp
http://www.dpreview.com/reviews/canoneos10d/page18.asp

The different size and layout of the crops makes it a bit hard to judge
from these pages but I'd say not. I think the 1D-II at 1600 might be
fairly close to the 10D at 800, but not 400.

Bear in mind that DPreview used their newer testing method on the
1DMkII, so you can't compare the 10D & 1DMkII results directly. If you
look at the top of the first link you quoted, you'll find that they have
a notice there pointing that out.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

wrote:

Kibo informs me that Rob Davison stated
that:


wrote:

Kibo informs me that Alfred Molon stated
that:



des wrote:


According to Canon press release images taken at ISO 1600 with the 20D
are approximately equal in noise to those taken with the EOS 10D model
at ISO 400. Pretty impressive if true..

How can that be since pixels are smaller (same CCD size with a higher
pixel count) ?


Better processing on the analog side, presumably, as with the 1DMkII.

I'm doubtful. If it were true I'd be really tempted. The 1D-II has a
much larger sensor and the same imaging chip as the 20D.


No, it's a different sensor.

I was talking about DIGIC-II. Somehow 'image processing' got turned
into 'Imaging'. Sorry.

I do realise the sensors are different. The 1D-II sensor being larger
should make for cleaner pixels was the gist of my argument.

http://www.dpreview.com/reviews/cano...kii/page18.asp
http://www.dpreview.com/reviews/canoneos10d/page18.asp

The different size and layout of the crops makes it a bit hard to judge

from these pages but I'd say not. I think the 1D-II at 1600 might be

fairly close to the 10D at 800, but not 400.


Bear in mind that DPreview used their newer testing method on the
1DMkII, so you can't compare the 10D & 1DMkII results directly. If you
look at the top of the first link you quoted, you'll find that they have
a notice there pointing that out.

:-)

Scrolling down I see the 10D is included in the luminance noise graph
on that page.

I'll be keen to see the test results for a production 20D but at the
moment I'm still just a bit doubtful about this claim.

My credit card hopes its all hyperbole for a start... :-)


Rob.
--