Choice of red filter for transformed measurements

To produce transformed measurements we observe in two colors.  When observing faint red stars,  V-Ic and V-Rc require shorter exposures than  B-V.

At least for APASS comp stars, where the Rc and Ic magnitudes are derived using the APASS sloan measurements, I've noticed that the Rc magnitudes in general have smaller uncertainties than Ic.  What other considerations should be included in the decision when picking the Rc or Ic filter?  Is there a preference for Ic or Rc for some classes of variable?  I'm thinking about absorption or emission lines in the red part of the spectrum.

Phil