AIND input format for data ingestion

[magland]

I'm looking into the AIND format option for inputting data into the pipeline. The readme states

aind: data ingestion used at AIND. The input folder must contain an ecephys subfolder which in turn includes an ecephys_clipped (clipped Open Ephys folder) and an ecephys_compressed (compressed traces with Zarr). In addition, JSON file following the aind-data-schema are parsed to create processing and NWB metadata.

Is there a way to get more details on that? What should be in the ecephys/ecephys_clipped and ecephys/ecephys_compressed folders? And why are they both needed? Isn't this a lot of duplicated data?

[alejoe91]

@magland sorry for my late response on this.

We're in the process of documenting the file-standards that we use and here's a draft of the ecephys standard:
https://github.com/AllenNeuralDynamics/aind-file-standards/blob/ephys-update/file_formats/ecephys.md