Question about LFP scaling and units in export_to_ibl_gui()

Hi,

while using export_to_ibl_gui() (SpikeInterface 0.102.3) to prepare data for the IBL ephys alignment GUI, I noticed what might be an inconsistency in the expected units of the LFP input data.

My understanding is:

• The LFP data used for calculating the LFP RMS values are expected in physical units (µV).
• The LFP data used for calculating the PSD values appear to be expected in ADC units and are converted to physical units (µV) within export_to_ibl_gui().

This made me wonder whether export_to_ibl_gui() is consistently handling scaling when computing the exported LFP features.

Could you clarify:

1. What units are expected for the input lfp_recording?
2. Are both RMS and PSD computed from scaled traces (return_scaled=True)?

I may be misunderstanding how the function interprets the LFP values, but I wanted to check whether there is a potential unit mismatch between the RMS and PSD exports.

Thanks!

Locate export_to_ibl_gui() and trace how lfp_recording is read in the RMS and PSD calculation paths, including any scaling options. Compare the documented input units with the values used for each exported feature; done means the unit behavior is clarified and any mismatch is covered by the relevant tests or documentation.