Fig 2. Data collection from ganglion cells in vivo.

(a) shows the smoothed time course of a single RGC fluorescence response to a 0.15 Hz luminance flicker stimulus. (b) shows the amplitude of the Fourier transform of the response time course, recovering the response peak at the stimulus frequency of 0.15 Hz (red dashed line). (c) shows the effect of the imaging light on cone-mediated response to stimuli in 62 foveal RGCs. The 488 nm imaging laser which excites GCaMP fluorescence can scatter to foveal cones and confound cone stimulation. If the imaging light is turned on without adaptation to any light, there is a quick rise and slow falloff in fluorescence due to cone responses feeding RGCs. However, if the 561 nm stimulus laser (used for drifting gratings) or LED white point (used for chromatic stimuli) is presented at a mean power of 2–2.5 μW during a short 30s adaptation period prior to imaging onset, then there is no corresponding effect of the imaging light confounding the fluorescence signal. In all recordings, an adaptation light preceded each recording to prevent this transient effect of scattered light from the imaging laser.