The Speckle pattern which Iam talking about is by considering the Fiber Specklegram sensors and when there is no external perturbation applied to the fiber. Here, fiber is MMF POF fiber.
Ideally speaking if you have complete control over the external perturbation sources like temperature gradients, acoustic vibrations, fixed fiber coiling and lengths, etc the speckle pattern should remain constant. Depending upon the sensitivity of your sensor (CCD) what it can resolve, will decide what kind of noise floor you will be looking at!!! If the sensor is of very high sensitivity it will peak up speckle pattern changes happening due to thermal gradients across the fiber length. If the fibre in use is in the range of few 10s of micron, it has a tendency to peak up acoustic vibrations.
All it depends upon what are the specs of your sensor and fiber and what kind of signal sensitivity you want to have. No matter what, you will see the noise, it is just one needs to know where the signal of interest is located in that noise!!!
I guess the must be some noise which is Poisson distribution fluctuation of number of carriers generated by photons.E.g. if you observe 10000 electrons signal in given pixel it will fundamentally fluctuate sample-to-sample with sigma = sqrt(10000) =100