Tbg, If you do agree with my statement above, then there is no way that recording engineers could "agree to one polarity and stick to it". I'm not absolutely certain that my assumption is correct (regarding the fact that on the recording side, selecting polarity is not possible due to multiple instruments, multiple mikes, the effects of mixers, room effects, etc), but that's what I wrote. True, the recording engineer could decide whether or not to flip a polarity switch, much as we do on the other end of the repro chain, but he's only altering the polarity of a bunch of mixed sources that in and of themselves most likely are of non-uniform polarity.