Does signal strength depend on exposure time and / or number of stacked subs?

[vlaiv]

33 minutes ago, andrew s said:

@vlaiv photon detection is a fundamentally quantum process. You can't think of a star emitting photons which travel to earth and are then detected on earth. This mixes the QM and classical views.

In QM there is a probabilty that photons are created at the star and given that a probability the will be detected (annihilated) here on earth. This is process is fundamentally probabilistic. Measurements over time will converge to the average  but will always retain some uncertainty for any real example. 

Regards Andrew 

Not sure what this relates to? Bose Einstein statistics of general idea that there is a signal of certain magnitude and that measurement yields numerical value that is polluted by noise?

If we are talking about signal and measurement, and if we have mathematical model that describes that process, why do you think it is not useful to think about it as if there were definite signal with definite photon rate, but it is the measurement that can't be done with absolute precision?

It's a bit like thinking about probability in propensity sense. We can say that there is a definite probability that dice will land certain number - that being 1/6 - although there is no finite number of measurements that will yield that value with certainty (in sense that adding more measurements can't disturb it / make it more precise).

We don't talk about that probability being dependent on number of measurements - we accept it as maybe fictional but in the same time for our purposes very real tendency with definite and precise numerical value associated.

[andrew s]

@vlaiv I was trying to point out that even with a noiseless perfect detector there will be variation in the measured arrival rate due to the physics of photon statistics. 

On top of this there will be noise added by the measurement process.

So on top of the 1/6 die face we have reading errors e.g. mistaking a 5 for a 6.

Regards Andrew 

[dph1nm]

On 20/11/2019 at 12:48, vlaiv said:

If we are talking about signal and measurement, and if we have mathematical model that describes that process, why do you think it is not useful to think about it as if there were definite signal with definite photon rate, but it is the measurement that can't be done with absolute precision?

 

Because this encourages people to think that it is their camera which is the cause of all the noise, and that by spending lots of money they will do better. In fact the signal from the source really does fluctuate and there is nothing you can do about this.

NIgelM

[vlaiv]

2 hours ago, dph1nm said:

Because this encourages people to think that it is their camera which is the cause of all the noise, and that by spending lots of money they will do better. In fact the signal from the source really does fluctuate and there is nothing you can do about this.

NIgelM

I did not think of that, you might be right, but I never said that measurement is solely dependent on kit used - in fact, you will be much better off throwing time at the problem rather than money

 

[andrew s]

23 minutes ago, vlaiv said:

I did not think of that, you might be right, but I never said that measurement is solely dependent on kit used - in fact, you will be much better off throwing time at the problem rather than money

 

As you expose a ccd it goes through 4 regimes. Initially it is read noise dominated. This is followed by photon/shot noise which is proportional to the square root of the signal. Next comes the region dominated by pixel response non uniformity (PRNU) which is proportional to the signal level. Finally we reach saturation where the ccd's (near) linear response to light no longer holds.

Once you reach the PRNU region extra signal does not improve the SNR you need high quality flats to remove the PRNU noise.

Regards Andrew