Dear Kotaro, 
It is probably due to the fact that your two half maps have the same
symmetry parameters. A more tedious way to solve it rigorously is to
divide the filament dataset into two unrelated halves and run two separate
refinements with slightly different starting symmetry parameters.
Convergence to a local minimum will be reached so that the two maps can be
aligned and used for a reliable FSC without further treatment of the
experimental maps. 
Good luck,
‹

Qiu-Xing Jiang, Ph.D.
Department Microbiology & Cell Science at UF IFAS
Electron Microscopy Faculty Director at UF ICBR
PO Box 110700 | 1355 Museum Drive, Rms 1298 or 1003, Gainesville, FL
32611-0700 USA
Biotech.ufl.edu <https://www.biotech.ufl.edu/> | Tel: 352.846.0953| Fax:
352.392.5933 |E-mail: [log in to unmask]






On 2/28/17, 11:17 PM, "Collaborative Computational Project in Electron
cryo-Microscopy on behalf of Kotaro Tanaka" <[log in to unmask] on
behalf of [log in to unmask]> wrote:

>Dear all,
>
>I noticed that the unmasked fsc of the two maps was too high in my case.
>
>The minimum of the unmasked FSC was 0.81, which is higher than the
>determination threshold of 0.8 for phase-randomization start resolution.
>
>I will try to lower the unmasked FSC in the very early iterations. Maybe
>by reducing Initial lowpass-filter value (now 20 A).
>
>Thank you.
>
>Best Regards,
>Kotaro Tanaka