If you're worried about sound quality and think you can hear a difference between 96k and 48k then why are you downsampling at all? Why not get a DAC that will handle your files? Theoretically, the decrease in quality would be much worse going from 96k to 48k no matter how evenly they divide. If I were someone who says they can hear a difference in sounds above what's audible to humans, I'd be more worried about that "quality" cut than whether my computer could properly do a math problem with something that's not evenly divisible. (I'm pretty sure it can).
FWIW, this "evenly divisible sample rate" paranoia is mostly left over from a time when there were unsophisticated resamplers and low-powered computers running them. Most modern resamplers will apply the right processing when needed.
Also FWIW, neither 96k nor 48k are bit rates, they are sample rates.