You could do what Nutul suggests and ensure that the sub and speakers are crossed over appropriately. Do this first.
After that, you could use REW to measure the system as a whole, calculate a convolution filter, and then upload that to CamillaDSP and set it to apply that convolution.
Note: Most of this is not nearly as straightforward as I have made it sound.
After that, you could use REW to measure the system as a whole, calculate a convolution filter, and then upload that to CamillaDSP and set it to apply that convolution.
Note: Most of this is not nearly as straightforward as I have made it sound.