The SPAD camera,47 as shown in Figure 2a, is a multipurposesingle-photon counting image sensor that utilizes a 32 × 64array of 30 μm diameter SPADs arranged in a 150 μm pitchsquare grid. Each pixel includes the necessary digitalelectronics for photon counting and features a very low noiselevel of 100 cps. Additionally, a microlens array optimizes thecollection efficiency, achieving a fill factor of up to 80% andhigh photon detection efficiency in the visible region, with amaximum of about 50% at around 410 nm. Three integratedcounters within each pixel can be enabled within a user-definedtemporal window (gate signal). By using two gate pulsesslightly shifted in time, a time resolution better than 200 ps canbe achieved by simply computing the difference between thetwo counters, effectively isolating the photons detected withinthe specific time interval defined by the shift. Because eachpixel has separate counters, we need to account for variationsin the gate pulse arrival time. In Figure 2b, we show thedifference in arrival time across the chip, which can be an up to300 ps delay between the opposite edges. This variation mustbe corrected to retrieve the full 200 ps time resolution whenperforming spatiotemporal measurements (Figure 2b, inset).The optical setup is depicted in Figure 2c. Basically, it can beunderstood as an inverted wide-field fluorescence microscope,with the only difference being that a transmission mask isplaced one focal length away of lens L in the excitation path,which is imaged on the sample. By changing this mask we caneasily switch from a diffraction-limited point excitation (using a20 μm pinhole) to a high frequency sine-like excitation (byusing a 20 lines/mm amplitude diffraction grating), asrepresented in Figure 2d. This allows the achievement ofdifferent super-resolution strategies (as we will discuss later) inthe same setup and in a very convenient manner.A high NA (1.4, oil immersion) microscope objective is usedboth to excite and to collect the photoluminescence from thesample. A 90/10 (T/R) beamsplitter is used to separate theexcitation and detection paths instead of the more traditionaldichroic mirror in order to achieve high sample (wavelength)flexibility. A microscopy tube lens (Thorlabs TTL200-A) isused to generate an image of the sample on an intermediateplane. Because of the large pixel pitch of the SPAD array, anadditional telescope is required to achieve the requiredmagnification. For this purpose, a 10x (Olympus NA 0.25)microscope objective was used together with a 600 mmachromat lens, leading to an overall magnification of 2000x oran equivalent pixel size in the sample of 75 nm. As anillumination source, a supercontinuum laser (NKT PhotonicsSuperK Extreme) is used. A 5 nm bandpass filter is used toselect the desired excitation wavelength. Figure 2e representsthe pulse scheme used in the experiment. An electronic triggerpulse is obtained from the laser controller, which is then fed toan adaptive computing board (Red Pitaya, STEMlab 125-14),in charge of generating the two gate pulses. By electronically scanning these two delay pulses, a full movie can be obtained,showing the spatiotemporal dynamics of our sample. Noticethat the camera design circumvents the need for beam-scanning optics, typically required in conventional setups.Instead, it captures spatiotemporal data across the entire regionof interest simultaneously, providing a major advantage inspeed and greatly simplifying the experimental setup, which iscompletely free of moving parts