Until now, descriptions in the literature of capture systems for multi-channel visible-spectrum imaging (MVSI) have been dominated by a pair of designs. The first of these designs is extremely general. It acquires images through a large number of filtered channels each with a narrow spectral bandpass. Disadvantages are associated with low light throughput per channel and a torrent of image data. A second popular approach offers higher light throughput and lower data demands because it depends upon a limited number of wide-bandwidth channels chosen with prior knowledge of scene contents. Unfortunately, this second approach is customized to scene contents so it is not a general solution. Hybrid approaches are introduced here that combine the generality of a many-narrow-channels solution with the lower data demands of a customized, wide-band system.