Optimal way of doing it if forming correct response and slope in digital domain per each sub (and speakers of course) with PEQ filters, leveling them to each others and time align them. There are many ways to fry a chicken and depending on your needs you chose the most appropriate one for your use case that won't be too much expensive. So you can use a PC with either a multichannel I/O interface if you want or need to go balanced and have a very good ADC, mic amp's and such or you can go with deacent internal multichannel sound card but in that case you will be limited to desktop PC and very little of those with good ADC. General advantages are software flexibility (decoders, encoders, plugins, players, DAW's, convolver engine's, infinite PEQ's... pretty much anything ever written) and generally ability to use up to date and most advanced implementations (64 bit FP processing) and disadvantage is obviously having to deal with PC (complexity, maintenance...) and how they mostly ain't that silent or completely silent. Or you can go with all in one ADC-DSP-DAC in which case you are obviously limited to inbuilt functionality, capability and filter quality. MiniDSP Flex isn't a bad choice in this category right now all do filter quality isn't stelar and convolver import isn't as easy as it could/should be and PEQ limit is ten per channel which should be enough combined in most cases. It's also centralised volume control (preamp), only thing really lacking is loudness control (which you probably have on your current RME). Main advantage is obviously self stand operation (once you complete the setup) and that is applies to all available/supported inputs.
Best regards and have a nice time.
