Neil: discrepancies in 3.5 dataset. Excessive power in the peaks above the transfer frequency. Could look in static LISA response

Shane: maybe the barycentric waveforms have the wrong power

Michele: will provide spectra for that (AI)

Neil: bigger problems with LISACode, especially for A, E, and T levels and shape

Neil: would be nice to distribute the barycentric spectra

Michele: will mention at open telecon

Michele: at the open telecon: who’s doing what

Neil: ...any issues they’re having. Mostly remind them we’re doing this.

Neil: would be nice to have submission interface soon. To walk people through it... they should have freedom to return different combinations of parameters. In 3.4 and 3.5 only some combinations of pm and pd noise can be determined.

Michele: just as free textual input?

Neil: maybe let them define fields and give values for them; or choose between default and custom fields.

Neil: also problem with multimodality. What would be returned there? Return best estimate and downsampled Markov chains.

Michele: indeed, would be able to accommodate that in lisaXML, and produce marginalized plots, etc.

Neil: allow also free-upload capability.

Michele: will send announcement for open telecon

Neil: may take them through 1B submission interface, suggest how it’s going to change

Neil: also need to encourage early work