Incorrect Backward Eigenmode Calculation

[WillE9a]

Hi all,

I read up on some of the previous issues regarding incorrect backward mode calculations (issues #1773 and #1585 mainly). It seemed like #1585 was supposed to be resolved in v1.21(see #1593) but I still encounter the same problem when running some small tests with the newest version. I was hoping someone can either look into this or correct a mistake on my end if I am using something incorrectly. I've included the geometry, code, and resulting plots below

inf_wg_dipole.zip

To summarize, I ran a simple 2D simulation with a y-oriented dipole source in the middle of an infinite GaAs waveguide and was comparing the coupling efficiency of the forward(+) mode to the backward(-) mode (expecting those to be the same). However, that was not the case as I got the following plot:

I am not too concerned with the shape of the plots, only understanding why there is such a large disagreement between the forward and backward modes.

Worth mentioning: increasing the resolution pushes the (-) eigenmode coupling closer to the (+) result, but you have to use a pretty large resolution to get a perfect match (unnecessarily large for such a simple sim).

I would appreciate it if anyone is able to explain why this is occurring; Is it definitely a bug buried deep somewhere within meep or am I implementing something incorrectly(e.g, the method of setting up flux regions to record fluxes and get eigenmode coefficients)? Please let me know if I can provide anything else to help clarify the issue.

Thank you!

[oskooi]

What you are seeing has been previously reported in #1470.

[WillE9a]

Thank you for the quick response! That explains the issue I'm having

[stevengj]

Closing as a duplicate.