Carter Francis

@AbaPanu From my understanding the Broker M4 files are compressed but not chunked? @ericpre is that correct? Unfortunately that is the worst case scenario... Is there a computer with greater...

@AbaPanu the reason why things are dying is partially because `hyperspy` is trying to process things in multiple threads so there might be some duplication. I would start with something...

> Many thanks for adding this! @ericpre Just to note that you seem to have two methods to read mib files now, one `file_reader` and one `load_mib_data`. The `file_reader` is...

From @matkraj in https://github.com/pyxem/pyxem/issues/634 > Hi all, Is there any reason hyperspy/rosettasciio doesn't load Merlin data lazily as a default? Merlin data is easily memory mapped so it should not...

@matkraj I don't know if we want to set multiple defaults for if a signal is lazy or not lazy as it starts to get confusing. Memory mapping in my...

Just to give an extreme example I often have .mrc files that are something like 400x256x256x128x128 (time, x, y, kx, ky) in which case reading along the time axis is...

> Hi @CSSFrancis thanks for pointing this out for me. When I was working on GPU processing in past I preferred any geometry adjustments to be done after the data...

@uellue Thanks for the detailed information! It seems like LiberTEM is once again ahead of the curve :) > For 4D STEM and likely also spectroscopic data the [CSR format](https://libertem.github.io/LiberTEM/reference/dataset.html#raw-csr)...

> Is intensity here the "time over threshold" supported by some detectors, or is this a number of counted electrons from a time bin or similar? Having access to the...

Okay I think that I'm starting to get the scope of the problem but someone tell me if I am wrong: Data Acquisition and storage: ----------------- There are potentially 2...