Efficacy of a "loop-back" due to impedances

Hi, I was just pondering the sound card calibration system after noticing that the 'ripples' on the loop-back measurement from my audigy 2 zs notebook look quite similar to some of the ripples in my frequency response measurement of my speaker and it hit me: line impedance... since the microphone (preamp) to sound card will have a different impedance setup than the sound card loop-back setup, will that not affect the determined frequency response? would there be a better way than a loop-back cable to measure the calibration file? many of my measurements appear to have the same ripple response in the midrange (but inverted since it is 'subtracted... could this mean that the ripple responses may be less than accurate; one would assume that if you are subtracting out something that should always be there, you would not have any 'trends' between the measurements that look like what you are subtracting out...