EOS 20D Low Noise

Mitch Alsup wrote:

Technology moves forward. For example, is the 2 years since the 6.3 MP
with 1.6 crop was introduced, CMOS technology has advance to the point
where the transistors associated with each cell are now only 1/2 as big
as they used to be. In addition a resent Canon patent application shows
using 3 transistors for 2 cells whereas the previous sensor cell had two
transistors per sensor. Other advancements in semiconductors (borrowed
from DRAMs) allow a cell to retains its high 'full well capacity' even
as the footprint of the cell is reduced. Then outside the sensor, there
has been progress in lower noise amplifiers and A/Ds.

If camera senseor technology moves forward at the same rate semiconductor
technology moves forward, one can expect 10 MP at 1.6 crop in 2 to 2.5
years, and 14 MP at 1.6 crop by the end of the decade.

That's all fine, but then why doesn't the more advanced 2/3" CCD of the
8MP prosumer cameras have less noise than the 1/1.8" CCD of 5MP prosumer
cameras ?

Cell size is almost the same (2.8 micrometer for the 5MP CCD, 2.7 for
the 8MP one). But a test shows that the 8MP CCD has slightly higher
noise levels - the 8MP CCD has at ISO 50 the same noise as the 5MP CCD
at ISO 64. Or do you suggest that there has been only progress in the
CMOS area, but none in the CCD area ?
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html

Mitch Alsup wrote:

Technology moves forward. For example, is the 2 years since the 6.3 MP
with 1.6 crop was introduced, CMOS technology has advance to the point
where the transistors associated with each cell are now only 1/2 as big
as they used to be. In addition a resent Canon patent application shows
using 3 transistors for 2 cells whereas the previous sensor cell had two
transistors per sensor. Other advancements in semiconductors (borrowed
from DRAMs) allow a cell to retains its high 'full well capacity' even
as the footprint of the cell is reduced. Then outside the sensor, there
has been progress in lower noise amplifiers and A/Ds.

If camera senseor technology moves forward at the same rate semiconductor
technology moves forward, one can expect 10 MP at 1.6 crop in 2 to 2.5
years, and 14 MP at 1.6 crop by the end of the decade.

That's all fine, but then why doesn't the more advanced 2/3" CCD of the
8MP prosumer cameras have less noise than the 1/1.8" CCD of 5MP prosumer
cameras ?

Cell size is almost the same (2.8 micrometer for the 5MP CCD, 2.7 for
the 8MP one). But a test shows that the 8MP CCD has slightly higher
noise levels - the 8MP CCD has at ISO 50 the same noise as the 5MP CCD
at ISO 64. Or do you suggest that there has been only progress in the
CMOS area, but none in the CCD area ?
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html

"Alfred Molon" wrote in message
...
Mitch Alsup wrote:

Technology moves forward. For example, is the 2 years since the 6.3 MP
with 1.6 crop was introduced, CMOS technology has advance to the point
where the transistors associated with each cell are now only 1/2 as big
as they used to be. In addition a resent Canon patent application shows
using 3 transistors for 2 cells whereas the previous sensor cell had two
transistors per sensor. Other advancements in semiconductors (borrowed
from DRAMs) allow a cell to retains its high 'full well capacity' even
as the footprint of the cell is reduced. Then outside the sensor, there
has been progress in lower noise amplifiers and A/Ds.

If camera senseor technology moves forward at the same rate semiconductor
technology moves forward, one can expect 10 MP at 1.6 crop in 2 to 2.5
years, and 14 MP at 1.6 crop by the end of the decade.

That's all fine, but then why doesn't the more advanced 2/3" CCD of the
8MP prosumer cameras have less noise than the 1/1.8" CCD of 5MP prosumer
cameras ?

Cell size is almost the same (2.8 micrometer for the 5MP CCD, 2.7 for
the 8MP one). But a test shows that the 8MP CCD has slightly higher
noise levels - the 8MP CCD has at ISO 50 the same noise as the 5MP CCD
at ISO 64. Or do you suggest that there has been only progress in the
CMOS area, but none in the CCD area ?
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html


The CMOS has circuitry next to each sensor cell so there are room to improve
in this regard, shrinking the size of the electronic components, using
better micromirror etc. Still, there's a limit. Let's see what's the optimal
pitch size for CMOS sensors.

"Alfred Molon" wrote in message
...
Mitch Alsup wrote:

Technology moves forward. For example, is the 2 years since the 6.3 MP
with 1.6 crop was introduced, CMOS technology has advance to the point
where the transistors associated with each cell are now only 1/2 as big
as they used to be. In addition a resent Canon patent application shows
using 3 transistors for 2 cells whereas the previous sensor cell had two
transistors per sensor. Other advancements in semiconductors (borrowed
from DRAMs) allow a cell to retains its high 'full well capacity' even
as the footprint of the cell is reduced. Then outside the sensor, there
has been progress in lower noise amplifiers and A/Ds.

If camera senseor technology moves forward at the same rate semiconductor
technology moves forward, one can expect 10 MP at 1.6 crop in 2 to 2.5
years, and 14 MP at 1.6 crop by the end of the decade.

That's all fine, but then why doesn't the more advanced 2/3" CCD of the
8MP prosumer cameras have less noise than the 1/1.8" CCD of 5MP prosumer
cameras ?

Cell size is almost the same (2.8 micrometer for the 5MP CCD, 2.7 for
the 8MP one). But a test shows that the 8MP CCD has slightly higher
noise levels - the 8MP CCD has at ISO 50 the same noise as the 5MP CCD
at ISO 64. Or do you suggest that there has been only progress in the
CMOS area, but none in the CCD area ?
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html


The CMOS has circuitry next to each sensor cell so there are room to improve
in this regard, shrinking the size of the electronic components, using
better micromirror etc. Still, there's a limit. Let's see what's the optimal
pitch size for CMOS sensors.

Technology moves forward. For example, is the 2 years since the 6.3 MP
with 1.6 crop was introduced, CMOS technology has advance to the point
where the transistors associated with each cell are now only 1/2 as big
as they used to be. In addition a resent Canon patent application shows
using 3 transistors for 2 cells whereas the previous sensor cell had two
transistors per sensor. Other advancements in semiconductors (borrowed
from DRAMs) allow a cell to retains its high 'full well capacity' even
as the footprint of the cell is reduced. Then outside the sensor, there
has been progress in lower noise amplifiers and A/Ds.

Do you have a patent number and a site where I could download a copy
of the patent application. TIA

jpc

Technology moves forward. For example, is the 2 years since the 6.3 MP
with 1.6 crop was introduced, CMOS technology has advance to the point
where the transistors associated with each cell are now only 1/2 as big
as they used to be. In addition a resent Canon patent application shows
using 3 transistors for 2 cells whereas the previous sensor cell had two
transistors per sensor. Other advancements in semiconductors (borrowed
from DRAMs) allow a cell to retains its high 'full well capacity' even
as the footprint of the cell is reduced. Then outside the sensor, there
has been progress in lower noise amplifiers and A/Ds.

Do you have a patent number and a site where I could download a copy
of the patent application. TIA

jpc

(Mitch Alsup) writes:

If camera senseor technology moves forward at the same rate semiconductor
technology moves forward, one can expect 10 MP at 1.6 crop in 2 to 2.5
years, and 14 MP at 1.6 crop by the end of the decade.

But it's not going to. What people fail to grasp is that the rate of
development you can get out of near-analogue sensors with many many
levels is not going to follow the rate of development of digital
on/off memory devices.

Physical limits are looming a lot larger right now for imaging sensors
than they are for binary processing devices. Don't forget, also, that
the development in process technology that you've seen improve other
electronic devices over the last two decades has already been applied
to current sensors.

B