McStas: Source_gen file-based input normalisation does not work as advertised

[willend]

The Source_gen component advertises that

Finally, a file with the flux as a
function of the wavelength [lambda(AA) flux(n/s/cm^2/st/AA)] may be used
with the 'flux_file' parameter. Format is 2 columns free text.

Additional distributions for the horizontal and vertical phase spaces
distributions (position-divergence) may be specified with the
'xdiv_file' and 'ydiv_file' parameters. Format is free text, requiring
a comment line '# xylimits: pos_min pos_max div_min div_max' to set
the axis of the distribution matrix. All these files may be generated using
standard monitors (better in McStas/PGPLOT format), e.g.:
Monitor_nD(options="auto lambda per cm2")
Monitor_nD(options="x hdiv, all auto")
Monitor_nD(options="y vdiv, all auto")

Multiple issues:

• flux(n/s/cm^2/st/AA) is rather a brilliance and further th flag flux_file_perAA allows changing this unit
• It seems that the "original" normalisation can only be retrieved with flux_file_perAA=1 plus using the Mean_ brilliance output from Brilliance_monitor and compatible xwidth/yheight/focus_xw/focus_yh
• The additional divergence maps from Monitor_nD(options="x hdiv, all auto") seem to mess up the normalisation

[willend]

For an example of a functional setup, see e.g. https://github.com/matteobsu/AMG-Beamlines-McStas/tree/main/BOA where

• https://github.com/matteobsu/AMG-Beamlines-McStas/blob/main/BOA/BOA_beamline_monitors.instr generates output in MCPL and Source_gen oriented formats
• https://github.com/matteobsu/AMG-Beamlines-McStas/blob/main/BOA/BOA_beamline_monitors_MCPL_in.instr picks up the MCPL
• https://github.com/matteobsu/AMG-Beamlines-McStas/blob/main/BOA/BOA_beamline_monitors_Synthetic.instr picks up output from Brilliance_monitor

[willend]

Postponing full resolution until after release of McStas 3.3 - work around by following the recommendations above.