It is not shown here, but when the mosaic is chosen to be bigger, the glitches span on a larger energy range.
Visualising the Ewald sphere is helpful to understand why. Glitches occur when the Ewald sphere intersects with one or more point(s) of the reciprocal lattice.
The reciprocal point occupies a larger space when the mosaic is bigger, therefore the intersection happens for a larger span when the reciprocal point has a bigger mosaic.
The Ewald sphere of radius `k=2*pi/λ` grows with k (i.e. the energy). When the energy increases, the radius increases and so does the sphere's surface. This larger surface is more likely to intersect with reciprocal points. This is why more glitches can be seen for higher energies.
Looking at the McXtrace data, the positions of the dips are correct. The intensities aren't.
Simulating the same exact beamline might solve the latter issue and the intensities might match more.
Also, some procedures done above might differ from the ones ssrl-glitch db has done. See the questions section and the detuning part.
When the source is monochromatic, there is one Ewald sphere that intersects with the reciprocal lattice.
When the source is white, there are many Ewald sphere's (because there are many energies, therefore many sphere's of different radiuses) that intersect with the reciprocal lattice.
When the source is monochromatic, there are a lot more glitches with dips of higher absolute amplitude than when the source is polychromatic.
One possible explanation (that might be false) is the following; the cc selects one energy at a certain angle but it is not an infinite dirac peak for that single energy, rather it is a (very) narrow gaussian like function, one can look at the rocking curves in the detuning folder as an example.
This means that for many of the glitches, it is possible that some rays with an angle(so an energy) close to the energy intended to be selected pass through the CC and hit the monitors, making the glitch of the energy intended to be selected, lower in absolute amplitude or even disapear.
Only glitches that have quite a few energies close by that are also glitches retain high absolute amplitudes because the energy there is stolen and therefore there is a decrease in rays transmitted to the monitors.
When the source is monochromatic this problem does not exist because the energy the cc selects is the energy the source sends (modulo a very small dE of 1eV). Therefore the dips in intensity can have high absolute amplitudes for all glitches because only the rays of that energy go and hit the monitor(or don't as much, leading to a glitch).
Here are two pictures illustrating the above explanation:
###### White light :

- McXtrace finds a glitch for an energy where ssrl glitch db and the theory glitch python script don't.
- A theory glitch is found for an energy but neither ssrl glitch db nor McXtrace do.
See annex for pictures illustrating the above.
The most interesting case would be if McXtrace and ssrl glitch db find a glitch for an energy but the theory glitch python script does not.
Before continuing with interpreting the data, making sure the following questions are answered to lift some doubts and implementing the improvements might be a good idea.