search

oinanoanalysis / h5oina

Oxford Instruments NanoAnalysis HDF5 File Specification
15 stars 8 forks source link

Cannot determine binning factor from some EBSD "Camera Binning Mode" dataset strings #5

Open hakonanes opened 1 year ago

hakonanes commented 1 year ago

Dear H5OINA developers,

I'm unable to determine the binning factor from some EBSD "Camera Binning Mode" dataset strings, like "Speed 2 (156x128 px)". Can I determine it in some other way from such files?

I believe patterns in files with these types of binning modes are obtained on Symmetry Sx detectors (x = 1-3) where a part of the detector is unused (say the upper 96 rows -> (1344 - 96) / 8 = 156 px). Can I assume that these detectors always have the same number of unbinned pixels (say 1344x1024), so that I can reliably calculate the binning factor from "Speed 2 (156x128 px)"?

ppinard commented 1 year ago

Hi @hakonanes, I forwarded your question to our EBSD expert. I'll let know when I get a reply.

hakonanes commented 1 year ago

Fantastic, thank you @ppinard. I'm asking because I think better reading of file metadata in third-party software and libraries will benefit H5OINA users who value interoperability.

k8macarthur commented 1 year ago

Hi Hakon, here is a response below from our EBSD team. I will look into removing the word 'binning' from the other parameter and leaving it as 'Camera mode' to make it less confusing.

Symmetry’s resolution mode is 1244x1024 resolution, this is effectively the unbinned pattern resolution. The lowest resolution modes (Speed 2, 3, 4) are 156x128 resolution, which, as Hakon probably noticed doesn’t quite divide in to the full 1244x1024 resolution. In this case, we do a little extra calculation which allows us to bin a pattern by a factor of 8 (1244/8 = 155.5) and get 156 columns. To calculate binning factor it’s easiest to just use the pattern height from the camera mode, which should always divide in to the full unbinned resolution: (Binning factor) = 1024 / (pattern height). The only exception to this is the Speed 3 camera mode (156x88), which is basically the Speed 2 camera mode with 20 rows from the top and bottom of the pattern discarded. So still binning factor 8.

hakonanes commented 1 year ago

Hi @k8macarthur, thank you, @ppinard and the EBSD team for the detailed answer. This is all the information I need, as far as I can tell. Feel free to close this issue.

I will look into removing the word 'binning' from the other parameter and leaving it as 'Camera mode' to make it less confusing.

That would be nice, as having "binning" in the name sets an expectation of a self-contained conversion factor.