Suddenly, several hours ago, 60 minutes bars of several pairs got wrong and had a wild values like 100000 or -7800 . Restarting NT on the same data gave another wild values but at another time and not for the same pairs. 59 min bars were initially ok, but afterwards were disturbed either.
Upgrading from Beta 20 to Beta 22 did not help.
Definition of data series as 3600 seconds instead of 60 minutes helped and the bars are clean.
It seems something got wrong with 1 minute data computation.
I'm in doubt whether the bug is reproducible, as every time a wrong bar appeared at another time and had another wild value.
I had a similar problem with Renko bars in earlier betas and the corrected bug was 0 volume bars not filtered out.
A question: Is it possible to cause NT to recompute minute values ? (of course except of exporting a data to history text files and reimporting).

Comment