Does taking light images continuously change the data. (take light data at various inverse frame rates, for a while. Compare the first image against the latest so see systemic effect)
If yes
Does taking darks fix this effect (Perform above experiment and punctuate with darks)
We need to write out our exact plan for how to optimize: