Extended Data Fig. 2: Microscope calibration.
a, Theoretical autocorrelation function (ACF) for a sample of 10 particles diffusing in 2D through a point spread function (PSF) with a waist of 0.3 μm and D = 0.1 μm2/s. The pixel dwell time should allow for an accurate recording of the fluctuations. A general guideline is for the dwell time to be ~10 times smaller than the decay time of the diffusing species. In this case, the decay time is of the order of 300 ms, so any pixel dwell time smaller than 1 ms would be a very safe choice. The characteristic dwell time used in our temporal brightness acquisitions (2.4 μs) is way below this value. However, because the associated frame time is of the order of 640 ms on the Leica microscope, we felt that this was the best compromise between an acceptable photon collection and a not-too-slow acquisition time (about >1 min for 100 frames). b, Apparent brightness B versus intensity scatter plot originating from a movie (256 × 256 pixels, 100 frames) of a homogeneous mixture of Alexa488 imaged in a 90% (wt/wt) glycerol/water solution, for increasing values of the laser power. c, Change of brightness (fold change) as a function of the increase in intensity (fold change). As the mean pixel intensity increases (almost linearly with the laser power), the mean brightness scales proportionally. The linear fit (constrained to 0), has a slope of 0.9, indicating that the increase in intensity is matched by a proportional increase in brightness, as expected. d, Example of a dark count histogram for the Leica SP8 HyD photon-counting detector and for the analog PMT, the latter superposed to Gaussian + exponential fits (black dashed lines) to determine calibration parameters (see Box 2).
