Re: ISO 200000 ?

Kennedy McEwen wrote:

In article <43D2025F.4090203@xxxxxxxxx>, "Roger N. Clark (change username to rnclark)" <username@xxxxxxxxx> writes 
 What he failed to clarify is that in none of the references,
do people actually claim a detection with one photon.
For good reason - I never claimed detection of an object with a single photon either! 
The Poisson noise limit is NOT "Roger's limit."  It is fundamental
math and physics.


Kennedy,
Let's step back and try and cool down.  I think what might be the
crux of our problems are were are interpreting each other's
statements differently.  But maybe we are saying the same thing.

I never specified a "limit" to useful information.  I originally
stepped in and gave an example of detecting 10 photons saying that
Poisson statistics say it would give you S/N of 3.2.  Any limit
I was talking about was the Poisson statistics limit, not any hard
lower limit.

You challenged my 10 photons and said try it with less.  After multiple
back and forth discussions, let me interpret where I think
we are each coming from.

I think you are coming from a point of view of photons/pixel/frame.
I was thinking total photons, however you got there, whether you
get that from one frame averaging individual pixels, or from averaging
multiple frames.  For example, If you said 0.5 photons per pixel per
frame and you did one frame and used 100 pixels to detect an object,
I would say you got 50 photons.  Correct?  Then I would say
the best signal to noise ratio you could achieve is square root 50.
Correct?  That is what I meant by the limit to signal to noise:
the Poisson statistics limit.  I specified no other limit,
at least I can not find a post specifying another limit, nor
did I intend one.

You keep repeating the < 1 photon per pixel per frame, but I have
not understood from what you have written, how many total photons
you would consider a detection.  Whether it be multiple pixels
from one frame versus multiple frames does not matter, simply total
the number of photons from the subject: #pixels * #frames * #photons
per pixel.  From your original challenge to me of do it with less
than 10 photons, I took it to mean you want to make detections
with fewer than 10 TOTAL photons.

Is this a correct?

So, let's try and keep the same definitions.  If you are
comfortable with #pixels * #frames * #photons/pixel = total photons,
then let's discuss what are reasonable limits for detection.
I suppose it would be subject dependent.  If you want to
identify a military target, I would hope the S/N would be
high before shooting, thus many total photons detected.
If a science claim of detecting some galaxy at the edge of
the universe where no one will die, then perhaps a 1 or 2 standard
deviation detection is acceptable for publication.


have made it perfectly clear throughout that the issue is that your detection threshold of 1 photon/pixel/frame is not always correct. 

I never said there was a 1 photon/pixel/frame limit.

In fact there is only one condition where it is correct, and that is where the detection is achieved in 1 pixel and 1 frame! In almost all cases the object being detected covers several, even many pixels in the frame whilst in some cases many frames can also be used. Both spatial and temporal integration lower the threshold of SNR per pixel per frame required for detection. 

I agree.  And that is where the total pixels:
#pixels * #frames * #photons/pixel >> 1.

 If you really had 1 photon, and a perfect
system with no noise, you would have a 50% chance of correctly
saying when you detected that photon.
Yes, assuming you mean "no excess noise", only photon noise, not "no noise" at all. 

I agree.

 No scientific journal will accept
a paper where you say you detected something but it was 50-50 chance.
I would hope no police/military decision to shoot was decided
on 50-50 chance.  Oops-better not go there...
That is precisely the point - 1 photon/pixel/frame is not a 50:50 chance, not even when the sensor excess noise itself is the same level. Often if not always, excluding a few exceptions, the object being detected subtends many pixels and can be viewed over many frames. Both criteria stack the probabilities much better than 50:50. 

Again total photons = #pixels * #frames * #photons/pixel >> 1.
So I agree.

You may well be stacking many frames to obtain images of the rings, but I hope that JPL are doing better than just integrating frames. You know the general shape of the rings, their spatial extent, and should therefore be able to use that to aid the detection process. I suspect that the need for integration is that they want to see the structure of the rings, not merely detect their position or presence. 

No, both position and extent.  It is a hard problem.

With lower read noise cameras, like the 20d and 20Da having about 3 times
less read noise, and working in darker skies, are getting much much fainter,
and thus working near and below the 1 photon/pixel/frame level.

No, they are working near and below 1 phot/pixel/final image - you have already stated that the noise is 3phot/pixel/frame. 

No.  Read noise is electrons not photons.
Incident light * QE < 1 photon/pixel/frame.  The 3 electron read noise
adds to the 1 electron noise due to the 1 photon.

So, in summary, 1 photon per pixel per frame IS useful, if you
sum many pixels to get many photons.  Whether you do that with
pixel summing and one frame or multiple frames is irrelevant, but
you need multiple photons, and in all cases, your maximum signal-to-noise
ratio is the square root of the number of pixels you counted.

Yes, that is Poisson statistics, but it isn't the limit that you stated in previous posts. 

I hope this clears things up.  The above is the same limit
I stated in previous posts.  The only limit I stated was
noise due to Poisson statistics of total counted photons.

I will try your test of imaging an overcast night scene away from
city lights with a DSLR.  This could take a while.  Where I live,
Denver area, when there is heavy overcast at night away from the city
it is generally snowing and I probably do not want to be out in
such a storm.  Maybe someone else who has something like a 20D
and a 50mm f/1.8 or 1.4 lens can try it at ISO 3200 recorded
in raw mode, and stretch the result in the raw converter (before
conversion), then refine in photoshop to bring out details
and post it for us.

Roger
.

Relevant Pages

  • Re: I dont understand EPR
    ... which says that paired photons ... thing (lambda) that determines the results in individual ... polarization of the photons incident on the polarizers ... rates of coincidental detection, ie., that lambda is ...
    (sci.physics.research)
  • Re: ISO 200000 ?
    ... I was thinking total photons, however you got there, whether you get that from one frame averaging individual pixels, or from averaging multiple frames. ...
    (rec.photo.digital)
  • Re: Einsteins energy formula E=mc^2 is WRONG!
    ... Not so for photons with wavelengths of 9E23 km, ... The actual existence and detection of a very low frequency photon, ... that determination by changing frames to a frame in which the light has ... One cycle per second is far from one cycle per 100 billion years. ...
    (sci.physics.relativity)
  • Re: twins versus quanta collapse
    ... Along path A a string of entangled photons is detected ... After many attempts the statistic fits the wave function ... the detection of one of the two should not ... entanglement has fixed the two from the begin. ...
    (sci.physics.relativity)
  • Re: Einsteins energy formula E=mc^2 is WRONG!
    ... Not so for photons with wavelengths of 9E23 km, ... The actual existence and detection of a very low frequency photon, ... that determination by changing frames to a frame in which the light has ... One cycle per second is far from one cycle per 100 billion years. ...
    (sci.physics.relativity)