AC/DC - For Those About to Rock!

My original plan was to employ DC voltage for heating the tubes, but keep prototyping simple by using AC.  DC heater voltage is used mainly because it should reduce the amount of ripple induced into the DC rails and signal path, especially for Directly Heated Tubes like the 6B4G.  That should reduce the audible "hum".  Today I tried DC (added rectification and smoothing) for both the driver and output tubes.  I was surprised that for the driver tube, DC seemed to make absolutely no audible difference to the small hum I hear, and no measurable difference to the measured ripple.  DC on the output tubes, actually increased the hum.  I used one very one high quality 4500uF cap and a low EMF 15,000uF cap, so the DC was about as smooth as it could reasonably be.  With DC, one lead of the heater is grounded while the other is at 6.3v.  By contrast, using AC, both leads are connected to the transformer windings with the center tap at ground, so essentially, each heater lead is at 3.15v.  All I can think is that the symmetry of the AC is preferable even if it's oscillating.  The hum really isn't that bad, so for now, I'm gonna stick with AC heaters.