[Wien] Optics case.mat_diag instead of case.outmat

[yukihiro_okuno at fujifilm.co.jp]

Dear Dr Steltsov and Wien Users.

Thank you for your advice.
I will do your advice.
I thought the band index (lower and upper) is the index which I use in
calculating the dielectric function,
why  is  such a high index number of band needed? If I need such high
number of upper bandindex, ( I thought
this is corresponds to taking into account the many conduction bands)
should we take more big value for
EMAX in case,inopt ?


And the optics generate case.mat_diag instead of case.outmat. I tried to
find about case.mag_diag in the Wien2K
manual, but I couldn't find it.
Why such a situation occurs ?

I want to know the reason, are there somebody who know it?

Sincerely.

Yukihiro Okuno



                                                                                                                                     
                      Serg Streltsov                                                                                                 
                      <streltsov at optics.i         宛先:    Yukihiro Okuno/FUJIFILM at FUJIFILM                                          
                      mp.uran.ru>                 cc:                                                                                
                                                  件名:    Wien2k: Optics                                                            
                      2005/04/14 23:27                                                                                               
                                                                                                                                     
                                                                                                                                     




Dear Dr. Okuno,

I cannot reply on all your questions - there to much of them. I've just got
the suggestion to increase the UPPER BANDINDEX (to 100 for instance). That
easily could be that you just do not take the corresponding bands for the
optical calculations. This is especially important for example then you
have
to calculate big supercells and many slightly different bands as a result.
You can try to increase Emax for matrix elements calculation as well.

Sincerely yours,
Streltsov Sergey
------------------------------------
Institute of Metal Physics
18, Sofia Kovalevskaya St.
GSP-170 Ekaterinburg 620219
Russia
Tel: +7(343)3783665
Fax: +7(343)3745244