A note on the convex infimum convolution inequality