Azimuth 2020

MiJostyn.....not in all cases If the pivot of the arm is at a different angle than is the headshell..... Cleeds is right. Bad arm design....every warp of the record would change the azimuth (as well as the VTA, and vtf)

There is no RIGHT or WRONG setting or answer, only the best scenario for your situation and setup. You need to decide do you want the most accurate channel balance or the most accurate crosstalk measurement? 

Then sit back and enjoy.

That's not really true. "Crosstalk" is not a matter of being accurate. You can maximize the minimum crosstalk, i.e. lowest in each direction, or maximize the total cross talk. Neither is going to optimize azimuth which is accurate orientation of the stylus w.r.t. the groove.
Optimizing the left and right channel frequency response at high frequencies, i.e. >20KHz, appears to be the best measure of accurate azimuth. This is not the same as low frequency channel balance.

There is no RIGHT or WRONG setting or answer.

Well what you mention is what I meant: maximize the minimum, minimize the numbers, however you want to describe it.
Crosstalk always falls under the discussion of azimuth, hence the title of this thread and it was mentioned, so people are talking about it.
The point is, you can adjust one or the other till you die as the setting will never be 100% perfect, considering you have zero control over the playing surface and how a record was mixed, mastered, cut and pressed, even test records vary.

I have been using TTs for about 40yrs and have had many tables and different carts including some 3x as much as what I have now. Back in the day setting azimuth by eye and using a volt meter never got me the sonic results I get now with the Foz. Once plugged in takes less than 5 min to get all readings.
My soundstage is beyond wall to wall, I have almost no surface noise and wide dynamics and high resolution.

I'm very happy! 

Control of cutting heads is pretty accurate, and pressing would not impart angle accuracy. It would be highly unlikely the grooves would be out of alignment by more than say 0.05 degrees, and likely less.  I don't know all the inner workings of the Foz, but based on the stated method of crosstalk, I can tell you it is not optimizing azimuth, which means it is not optimizing what the needle can extract from the record.