Frequently Asked Questions

Why is Takin's sample rocking angle (a3) different than at instrument X?
Different instrument control systems have different conventions on where the sample angle's zero position is. When using different programs, you should either only use the deltas of this angle or find out the offset -- usually a multiple of π/2 -- between the systems.
Why does the resolution convolution simulation or fit not start?
This might be due to a configuration error or -- in the case of Python S(Q, E) models -- a syntax error in the Python script file. As the convolution tool delegates the actual work to sub-processes, it does not directly show errors from these processes. Instead any errors can be found in the log file (in Takin's main window select "Help" -> "Log..." from the menu) or on the console if Takin is launched from the command line.
Will fitting/data treatment feature X be added to the Scan Viewer?
The Scan Viewer is a lightweight tool used to have a quick look at scan data files while featuring some very basic fitting. Adding advanced features would make it more difficult to use for the simple tasks it was designed for. Advanced data treatment will instead be available in the separate Scan Browser of which a preview version is already available in Takin.

Notes

Scan files from PSI that were created by their new Nicos system seem to set "PARAM: FX=1" for kf=const mode. For the moment you have to manually change this in the file to read "PARAM: FX=2" for Takin to correctly interpret it.
Nexus files from ILL can have wrong scattering planes, you have to manually correct them for the moment.
The Eckold-Sobolev resolution method does not directly consider the sample geometry in its resolution matrix calculation. It instead requires a secondary convolution over the sample extents.