-
Notifications
You must be signed in to change notification settings - Fork 2
Time resolved experiments (with timing tool)
When the delay between the excitation laser and the XFEL pulse is in the femtosecond or picosecond region, the timing jitter becomes significant. In this case, you have to analyse the actual delay by the timing monitor tool.
Cheetah can sort images as described in Time resolved experiments (without timing tool), but you might want to bin images further. The actual method to get t0 (absolute time zero) and relative jitter by the timing monitor depends on your experiment and is not discussed here. We assume that you know which bin each image (tag) should go to.
(See also Dealing with unusual geometry and mask if you have big shadow from the laser beam stop)
With partialator
, you can specify
custom data splitting
to merge images in a stream file into several .hkl
files.
Create a custom split file bins.lst
as follows:
run266714-1.h5 tag-127921948// 100FS
run266714-1.h5 tag-127921950// 100FS
run266714-1.h5 tag-127921960// 200FS
run266714-1.h5 tag-127921968// 100FS
run266714-1.h5 tag-127921974// 200FS
...
Here, 100FS
, 200FS
are the names of the bins.
Run partialator
with --custom-split
option:
partialator --custom-split=bins.lst -i 266711-266721-dirax.stream -o 266711-266721-dirax.hkl --model=unity -j 14 -n 4 --output-every-cycle
This will create .hkl
files for each bin and for each iteration, which can be converted to MTZ files by create-mtz
script. Of course you might want to play with options like --model=scsphere
(partiality correction, but unstable), --push-res=1.0
, etc.
If you use process_hkl
, you have to make separate stream files for each
bin by running indexamajig
several times with different input lists.
In Running-CrystFEL, the input file contained file names alone. Here,
we list event names as well.
100fs.lst:
run266714-1.h5 tag-127921948//
run266714-1.h5 tag-127921950//
run266714-1.h5 tag-127921968//
...
200fs.lst:
run266714-1.h5 tag-127921960//
run266714-1.h5 tag-127921974//
...
Next, index and integrate separately to produce stream files for each bin.
indexamajig -i 100fs.lst -o 100fs.stream ...
indexamajig -i 200fs.lst -o 200fs.stream ...
Stream files can be processed by process_hkl
as usual.