dynamic range and thermal noise

dear ng,

I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?

Best regards for your input!
Marc Wossner

On 7 Okt., 15:38, Scott W wrote:
Marc Wossner wrote:
dear ng,

I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?

Best regards for your input!
Marc Wossner

I believe you are really talking about dark current, which is very
temperature depended, not thermal noise. If so then yes for long
exposures dark current can become an issue.

Scott


Well I thought dark current translates into thermal noise so that both
are indeed the same.
Isn´t that correct?

Marc

Marc Wossner wrote:
On 7 Okt., 15:38, Scott W wrote:
Marc Wossner wrote:
dear ng,
I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?
Best regards for your input!
Marc Wossner
I believe you are really talking about dark current, which is very
temperature depended, not thermal noise. If so then yes for long
exposures dark current can become an issue.

Scott


Well I thought dark current translates into thermal noise so that both
are indeed the same.
Isn´t that correct?

Marc

Yes, the square root of the dark current is the thermal noise.
Dark current is generally low in modern cameras, a fraction
of an electron per second, so it takes a while for it
to become a factor. It is higher with higher temperatures,
like 100F it could be several/second. The simple solution
is take multiple short exposures and add them together.
Effectively, with this technique, dark current is not an issue.

Roger

Roger N. Clark (change username to rnclark) wrote:
snip
Yes, the square root of the dark current is the thermal noise.
Dark current is generally low in modern cameras, a fraction
of an electron per second, so it takes a while for it
to become a factor. It is higher with higher temperatures,
like 100F it could be several/second. The simple solution
is take multiple short exposures and add them together.
Effectively, with this technique, dark current is not an issue.

Roger

The first part is correct. The second isn't. Combining
several exposures reduces the noise according to the square
root of the number of images. Combining four photos
lowers the noise only by a factor of 2. To reduce noise by
a factor of 4, you would have to combine 16 images.

On Oct 7, 7:26 am, Marc Wossner wrote:
dear ng,

I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?

Best regards for your input!
Marc Wossner

Correct. There are several types of noise, including quantization
noise (from the A/D uncertainty), dark noise, photon noise, amplifier
noise, - that takes care of the common ones. Which one is the
limiting noise is harder to determine.

On Oct 7, 8:48 am, Marc Wossner wrote:
On 7 Okt., 15:38, Scott W wrote:



Marc Wossner wrote:
dear ng,

I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?

Best regards for your input!
Marc Wossner

I believe you are really talking about dark current, which is very
temperature depended, not thermal noise. If so then yes for long
exposures dark current can become an issue.

Scott

Well I thought dark current translates into thermal noise so that both
are indeed the same.
Isn´t that correct?

Marc

Not quite. Usually when we talk about dark current we mean the
current in the photodetector. But any electronic element can have
thermal noise in it.

Scott W wrote:
Roger N. Clark (change username to rnclark) wrote:
Marc Wossner wrote:
On 7 Okt., 15:38, Scott W wrote:
Marc Wossner wrote:
dear ng,
I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?
Best regards for your input!
Marc Wossner
I believe you are really talking about dark current, which is very
temperature depended, not thermal noise. If so then yes for long
exposures dark current can become an issue.

Scott


Well I thought dark current translates into thermal noise so that both
are indeed the same.
Isn´t that correct?

Marc

Yes, the square root of the dark current is the thermal noise.
Dark current is generally low in modern cameras, a fraction
of an electron per second, so it takes a while for it
to become a factor. It is higher with higher temperatures,
like 100F it could be several/second. The simple solution
is take multiple short exposures and add them together.
Effectively, with this technique, dark current is not an issue.

Roger

You might want to look at this
http://en.wikipedia.org/wiki/Thermal_noise

I have never heard anyone refer to dark current as thermal noise.

But forget that for the moment, in your solution of taking multiple
short exposures you will still end up with the same ratio of dark
current to signal current. What you can avoid is saturation due to dark
current, but you will get the same noise from the dark current
regardless of whether you break the one long exposure into a number of
smaller ones or not.

A reference directly discussing electronic sensors:
http://learn.hamamatsu.com/articles/ccdsnr.html
Scroll down to dark current:

"Dark noise arises from statistical variation in the
number of electrons thermally generated within the silicon
structure of the CCD, which is independent of photon-induced
signal, but highly dependent on device temperature. The
generation rate of thermal electrons at a given CCD
temperature is referred to as dark current. In similarity
to photon noise, dark noise follows a Poisson relationship
to dark current, and is equivalent to the square-root of
the number of thermal electrons generated within the
time. Cooling the CCD reduces the dark current dramatically,
and in practice, high-performance cameras are usually cooled
to a temperature at which dark current is negligible over a
typical exposure interval."

And reduction of dynamic range is esxactly why to take many short
exposures and add them For example, if you had a gain of 1 on
your ADC (1 electron = 1 DN or ADU), then your 12-bit ADC
saturates at about 4000 electrons (including a small bias),
so if you had 1 electron/sec dark current, a 1 hour exposure
leaves tou with little dynamic range. Twelve 5-minute
exposure would maintain most of your dynamic range. That is
what I was referring to as not an issue, as the OP was asking
about dynamic range. You are correct that the noise from dark
current is always a factor, and the only way to reduce it is
to work at colder temperatures (e.g. winter nights are better
for night sly imaging than summer nights).

Roger

Marvin wrote:
Roger N. Clark (change username to rnclark) wrote:
snip
Yes, the square root of the dark current is the thermal noise.
Dark current is generally low in modern cameras, a fraction
of an electron per second, so it takes a while for it
to become a factor. It is higher with higher temperatures,
like 100F it could be several/second. The simple solution
is take multiple short exposures and add them together.
Effectively, with this technique, dark current is not an issue.

Roger

The first part is correct. The second isn't. Combining several
exposures reduces the noise according to the square root of the number
of images. Combining four photos
lowers the noise only by a factor of 2. To reduce noise by a factor of
4, you would have to combine 16 images.

It is correct. Adding multiple shorter exposures maintains dynamic
range, which is what the OP asked. What you say is correct also, but
is addressing a different question. I should have added
"regarding dynamic range" at the end of my sentence.

More info on low light photography:
http://www.clarkvision.com/photoinfo...ht.photography

Roger

On Sun, 07 Oct 2007 19:13:58 -0600, "Roger N. Clark (change username to
rnclark)" wrote:

Scott W wrote:
Roger N. Clark (change username to rnclark) wrote:
Marc Wossner wrote:
On 7 Okt., 15:38, Scott W wrote:
Marc Wossner wrote:
dear ng,
I´m still into analog photography but have a keen interest in the
digital technique. As far as I understand it, the dynamic range of an
imaging sensor/camera combo is defined as the maximum signal divided
by the noise which is produced in the various stages. Does thermal
noise still play a vital role in this calculation (as far as longtime
exposures are concerned) or is it so successfully erased by the noise
reduction techniques that it doesn´t have to be be taken into account
in calculating dynamic range?
Best regards for your input!
Marc Wossner
I believe you are really talking about dark current, which is very
temperature depended, not thermal noise. If so then yes for long
exposures dark current can become an issue.

Scott


Well I thought dark current translates into thermal noise so that both
are indeed the same.
Isn´t that correct?

Marc

Yes, the square root of the dark current is the thermal noise.
Dark current is generally low in modern cameras, a fraction
of an electron per second, so it takes a while for it
to become a factor. It is higher with higher temperatures,
like 100F it could be several/second. The simple solution
is take multiple short exposures and add them together.
Effectively, with this technique, dark current is not an issue.

Roger

You might want to look at this
http://en.wikipedia.org/wiki/Thermal_noise

I have never heard anyone refer to dark current as thermal noise.

But forget that for the moment, in your solution of taking multiple
short exposures you will still end up with the same ratio of dark
current to signal current. What you can avoid is saturation due to dark
current, but you will get the same noise from the dark current
regardless of whether you break the one long exposure into a number of
smaller ones or not.

A reference directly discussing electronic sensors:
http://learn.hamamatsu.com/articles/ccdsnr.html
Scroll down to dark current:

"Dark noise arises from statistical variation in the
number of electrons thermally generated within the silicon
structure of the CCD, which is independent of photon-induced
signal, but highly dependent on device temperature. The
generation rate of thermal electrons at a given CCD
temperature is referred to as dark current. In similarity
to photon noise, dark noise follows a Poisson relationship
to dark current, and is equivalent to the square-root of
the number of thermal electrons generated within the
time. Cooling the CCD reduces the dark current dramatically,
and in practice, high-performance cameras are usually cooled
to a temperature at which dark current is negligible over a
typical exposure interval."

And reduction of dynamic range is esxactly why to take many short
exposures and add them For example, if you had a gain of 1 on
your ADC (1 electron = 1 DN or ADU), then your 12-bit ADC
saturates at about 4000 electrons (including a small bias),
so if you had 1 electron/sec dark current, a 1 hour exposure
leaves tou with little dynamic range. Twelve 5-minute
exposure would maintain most of your dynamic range. That is
what I was referring to as not an issue, as the OP was asking
about dynamic range. You are correct that the noise from dark
current is always a factor, and the only way to reduce it is
to work at colder temperatures (e.g. winter nights are better
for night sly imaging than summer nights).

Roger

Displaying the pitfalls of a slightly-above-average I.Q. without logic and
common sense. Smarter people may be able to build bigger bridges, but they just
as easily dig deeper ditches for themselves, without even realizing it. With no
way out than one day relying on someone smarter and wiser than them to save them
from themselves. They rarely allow that, they're too smart for that. :-)

On 7 Okt., 16:29, "Roger N. Clark (change username to rnclark)"
wrote:

Yes, the square root of the dark current is the thermal noise.
Dark current is generally low in modern cameras, a fraction
of an electron per second, so it takes a while for it
to become a factor. It is higher with higher temperatures,
like 100F it could be several/second.

Thanks for the clarification!

The simple solution is take multiple short exposures and add them together.
Effectively, with this technique, dark current is not an issue.

OK, but except those measures after taking the picture
is there something manufacturers build into their cameras to reduce
thermal noise?
Are there digital slrs that use active cooling or dark frame
subtraction as a standard?

Marc