Overlapping emission spectra from multiple fluorophores can be computationally “unmixed” post-acquisition. Spectral unmixing requires the ability to record emitted photons within narrow spectral bands (e.g. <10nm; from 450nm – 460nm), as well as knowledge of the full emission spectra of the fluorophores being separated. Although spectral scans can be performed many ways (e.g. in a fluorimeter), the most common application of spectral unmixing in light microscopy involves point scanning confocal microscopy. All emission filters are removed from the light path and then the emitted light at each voxel is passed through a prism or diffraction grating to separate out all of the colors. This “rainbow” is then projected onto an array of PMTs, physically separated at precise intervals to provide full coverage, without overlap, of the region of the spectrum being imaged (e.g. 32 PMTs each covering 10nm = 320nm, such as from 400nm to 720nm). Following acquisition, assuming the “fluorophore reference spectra” are available, software is used to unmix the signal generating individual images for each fluorophore. Additionally, spectral unmixing can be used to remove autofluorescence signal from samples, such as tissue specimens.