Видео

Hi Hemant. In most of the cases it is very straightforward. If you have challenging sample, start from 100x100 or 200x200 pixel map, dwell time around 10 usec, drift compensation on and optimized for periodic lattices. If the sample is too uniform, you might try to use separate area for drift compensation. If the sample is easy to damage, than you might need to play with dwell times and map sizes to see if you can find good balance. Please note, that you need at least Super-X to be able to get such maps and for alloys you probably need probe corrector.

@@TEM_guide thanks for the details I will try this and yes we have Cs correcter microscope.

Dear Olivier We could not reproduce the issue you report. Clearly integration works properly and Spectrum profile produces different noise pattern vs Intensity profile. Were you using profile with small number of pixel in integration width? If yes, than results might appear identical for datasets with very low statistics (differences just very tiny since integration is very small). May be the best is to contact via Velox@thermofisher and share the dataset. This way we can analyze where the issue is.

Thank you for your feedback. Yes we will be happy if more users aware about this channel..

We are happy to discuss it via email (in channel description).

Hi. In Velox you have couple of different export options. If you want to export raw Spectra per pixel from SI, you can use Lispix export. If you want to export actual values from quantified elemental maps, you can use export of Elemental maps and select tiff16bit or mrc. Both formats are exported without any normalization, i.e. in intensity per pixel is equal to at% of wt% in the map. In the AutoScript we do not yet support direct readability of emd Spectrum images. However, emd file is just HDF5 with maps saved as tiff16bit. I.e. you can use Python standard readers to read this data.