Problems in the Oasys treatment of the electron energy dispersion

[srio]

After some discussions with Jae-Hong Lim, Pohang Accelerator Laboratory, South Korea it has been revealed that the treatment of the energy dispersion has problems in Oasys widgets.

As a reminder, the electron energy dispersion significantly enlarges the undulator peaks and the divergences at high harmonics.
It is therefore important for calculating emissions at high harmonics. Its effect on the power load calculations maybe not very significant as most of the power is generated at the lower harmonics, and the total emitted power does not change.

As a summary:

1. ShadowOui Full undulator does not take into account the electron energy dispersion
2. ShadorOui SRW/Shadow undulator includes this parameter but does not seem to have an effect on the result.
3. xoppy undulator spectrum does include this effect when calculating the spectrum (with SRW calculations)
4. xoppy undulator power density does include this parameter (with SRW calculations) but does not seem to have an effect on the result.
5. xoppy undulator radiation does not include this parameter. It is included in the script, but it does not have any effect on the result (with SRW calculations).

[lucarebuffi]

SRW do take that into account, but only in a real ME simulation. I believe that the convolutional approach that I necessarily take in the source generation case, it is not accurate enough at high energies.