At 04:30 PM 9/10/2014, Tom Fine wrote:
>Isn't there some technique for full-track azimuth adjustment where you 
>take two feeds of the output, polarity-reverse one of them, input into two 
>console positions at unity gain and then adjust azimuth for greatest net null?

That technique, and many of the other techniques described in this thread, 
will give you the best phases match between the channels.  However, this is 
often *not* the azimuth adjustment that gives the best HF response.  When I 
do Plangent transfers, I can of course adjust the azimuth for best HF 
response, because I'm looking at a real time display of recovered bias -- 
but this means that the phase match between the channels is sometimes not 

What causes this?  Two things that I know of:  1) the gaps in the record 
head were slightly staggered; and 2 (probably more common) the recording 
was made with different bias on the 2 channels.  The recording is made at a 
point slightly past the trailing edge of the record head gap, and changing 
the bias changes this distance.  This is why bias should be adjusted before 
adjusting record head azimuth, or record head azimuth should be adjusted in 
sync playback mode, on machines that have that feature.

(And, BTW, my version of 20-20 hindsight is that record head azimuth should 
have always been fixed, and playback head azimuth should always be adjustable.)

Given that these days everyone is transferring to digital, and analog tape 
copies are very seldom made, phase alignment can easily be done with a 
digital tool, if it's not good enough when azimuth is set for best 
frequency response.

-- John Chester