any audio BJT low noise but very high voltage?

Simple relative way (done&tested already): maybe some forum friends here will appreciate it. Similar approaches can be found in many places in internet.
As said in beginning of the thread, I have made a small tiny coil (50um wire gauge, OD3mm Teflon former, R=10Ohm, L=150uH (air). Make sure the LR corner frequency is above 10KHz at least.
Except the coil, which has very fragile wire, no other special tools are needed.

A) for PLay
> Run reference tape: align the machine as usual for 0dB -10dB and -20dB.
Then put the tape away, and bring in touch the little pick-up coil to the Playhead's face/gap: use a reproducible and stable fixture (not by hand).
Inject signal in pick-up coil and run a discrete frequency sweep: record all measured values for each frequency. My suggestion is to measure the signal directly at the head's terminals (unsolder the connections to FE/preamp).
Make a table: reference tape vs. pick-up coil readings.
This concludes the relative calibration of the coil against Reference tape and against a given Play head type (same design, materials, etc).

> Apply the calibrated coils for new Play head, or new FE:
Simply place the pickup coil in the same position against the head.
Inject the signals in the pick-up coil: same values as in the calibration.
Re-align your FE to output the same dB as in the calibraton table.

b) for Rec
> calibrate your Play head+FE first
Run Rec to obtain same FE-output numbers (by tape).

Usually this is good enough. Having an absolute flux calibration is not intended for any user.
 
What about exchanging roles? Transmit on single tun and receive on tape head?
precisely so.
If this single turn fits the size of the gap, doubled by simulations to precisely evaluate the linkage at various frequency points: it becomes an absolute flux calibration.

But the reverse situation is still needed: measure Play head signal by Reference tape -> then inject same signal in the Play head and measure in the 1-wire pick-up coil - > then compare to simulations: all this bother is just for double check. Which is needed, to correctly conclude the 1-wire pick-up coil absolute flux calibration.
 
I am sorry, in Post#42 it was wrong to mention only the comparison to the Reference tape: that is a relative check, not an absolute one.

One more critical step is needed for absolute flux calibration: one must run the analytically/simulation computed signals which will produce the same gap flux. And, if lucky, the readings will fit the reference tape :)