Loopbacking and comparing samples in an ABX fashion (see FAQ, and
http://www.hydrogenaudio.org/forums/index....topic=12118&hl= ) after having matched the levels is a suboptimal solution. If the samples can be ABXed (to hear the difference in a blind test) it's OK, but if the ABX fails it can be blamed on the recording quality that masks the subtleties of the playback, or that compensate for the differences.
However it is the best solution. If you can hear it 100 % of the times, you might achieve the ABX test.
Using two computers plugged into the same ampli is more rigorous, but more difficult to setup (need two SB Lives, two computers... and after all some may argue that the power supply can make a difference !)
Use a track where the difference is obvious. Loopback the soundcard (set the record device to line in, not wave out

, play the track (in Winamp, for example), while you record it (with CoolEdit for example), then do the same with the other drivers, then in CoolEdit cut both recordings at the exact same place (sample wise if possible) to make short samples, then ask for their RMS levels, if they are different, make them match with a volume process (better processing both files to a common level to avoid reducing the quality of one of them only in the process), then the samples are ready for ABX.
Use a program like WinABX, Foobar2000, PCABX... there are links near the top of the FAQ.
You must choose a fixed number of trials in advance. 8 if the test is difficult and you don't have much time, 16 otherwise. Then if you get 7/8 or 13/16 at least, it's ok, there is an audible difference between them.
EDIT : Bleh already made measurments and showed a noticeable difference in sound quality. So it is possible that it is audible.