for everyone else, who`s interested:import arcpy
rows = arcpy.SearchCursor("observer_points")
row = rows.next()
while row:
arcpy.MakeFeatureLayer_management("path_to_file", "one-point-dataset", "ID = " + str(row.ID))
arcpy.Viewshed_3d("path_to_file", "one-point-dataset","path_to_view-raster"+str(row.ID)+"_v.tif", "", "CURVED_EARTH", "")
row = rows.next()
yet this is about 3secs per point on a thinkpad Core2Duo 2.53Ghz 8GB Ram, Intel SSDand about 12secs per point on a pentium quad core 2.4Ghz, 4Gb Ram, 7200RPM Harddrive.Nevertheless it is taking to much time... 😞