Silicon photovoltaic solar cells generally have a black or blue appearance that makes them aesthetically very different from traditional red roofs that either comprise burned-clay tiles or composite-material shingles. Rooftop solar cells may become more acceptable if they are similar in appearance to traditional roofs. This objective requires that the red part (620-700 nm wavelength) of the incoming solar spectrum be reflected so that it becomes unavailable for photovoltaic generation of electricity. Complete reflection of red photons would result in the reduction of useful solar photons (300- 1200 nm wavelength) by 12.5%. Calculations show that the optical short-circuit density will then decline by: 17% for 100-μm-thick crystalline-silicon solar cells, 20-22% for triple-junction tandem thin-film solar cells of amorphous silicon, 15-16% for 2.2-μm-thick CIGS solar cells, and 16-20% for ultrathin CIGS solar cells. On average, the efficiency of a typical solar cell will have to be multiplied by a factor of 0.8 if all red photons were reflected. This reduction in efficiency can be offset by wider adoption of rooftop solar cells. Red-rejection filters can be made of particulate composite materials containing, say, silica nanospheres. Typically, the solar cells will be iridescent then, which may not be aesthetically pleasing to many. Non-iridescent red-rejection filters can be fabricated by upscaling the linear dimensions of biomimetic filters nano-imprinted to reproduce the Morpho blue, this possibility being guaranteed by the scale invariance of the Maxwell equations and the weak dispersion of the refractive indexes of numerous polymers in the visible spectral regime. Non-uniformly red rooftop solar cells would also become feasible.