Interpolation (?) of Phoenix spectra in the UV

We ran into a strange effect when trying to simulate a Phoenix spectrum in the mid-UV (120-220nm). Here only high effective temperatures have flux, so a model spectrum Phoenix(t_eff=12_000) was called to test the spectrograph model. The output shows not a continuous spectrum, but the only wavelength values with flux are on a regular grid - the stellar spectrum is modulated by an etalon-like effect. Not sure if this is coming from the wavelength grid of the synthetic spectra or a bug in the interpolation, haven't had a chance to look into it myself.

The image shows two traces, one with the Phoenix spectrum, and second with a Blackbody(temperature=12_000) spectrum as comparison.