More

Standard treatment of BRISP raw data

The standard treatment of the BRISP data allows for transforming the 3D raw data (x_detector, y_detector, time channels) to an intensity I(Q,ω) which is closely related to the dynamical structure factor S(Q,ω), the difference mainly lying in additional corrections that may be needed such as the standard (trivial) empty cell subtraction and, if needed, the multiple scattering subtraction. 
The procedure used on BRISP transforms directly the intensities measured in the individual pixels to I(q,w) without any radial integration or energy rebinning. 

The typical actions carried on the BRISP data can be found in the brisp_pix2sqw.prox, and are below listed: 

1. w1=rdrun(24420) 
   Short measure of direct beam (numor 24420 in the example) made by removing the beam stop and inserting an attenuator on the beam. 
2. C0=brisp_center(W1,Rb_stop=8,mask_in='msk128',mask_out=W60) 
   Calculate the beam centre (C0) and create a detector mask (W60) composed by the standard mask (msk128) combined with a circle representing the direct beam (having a radius of 8 eight pixels in this example) and centred in C0. 
3. 3) w2=rdsum(24421,24433) & w2=brisp_par(w2,lam=0.988,dst=4.) 
   Sample scans may be summed, and the right parameters are written in the file. 
   Three choices possible for lambda (lam): PG002: 0.988, Ang PG004: 1.977 Ang, Cu111=1.28 Ang. 
   Sample-detector distances (dst) allowed on BRISP range between 2 and 6 m. 
4. w52=total(w2,3)*(1-w60); 
   Possibility of looking at 2D (time integrated projection), just for visual purpose. 
5. w12=brisp_rad(W2,XY0=C0,INCR=6,MASK=w60) 
   Radial integration providing I(theta,channels). This quantity is calculated by incrementing the angular step of integration by a constant number of pixels (in this example the choice of 6 pixels combined with a dst=4m give an angular step of ~ 1o). 
   The calculation of this quantity allows to evaluate the time channel range where the signal is only background for further correction (see next step). 
6. 6) w4=raw_bck(w2,bck_ch=[1100,1600],mask=w60,kernel=[9,6]) 
   Calculate the ambient background in each detector pixel (w4 has the same dimension of raw data) by: 
   (i) propagating the counts measured in a channels range where only background contributes (from 1100 to 1600 in this example) 
   (ii) taking an average in a pixels window (9 times 6 pixels in the example) 
7. 7) w22=sqw_pix(w2,XY0=C0,MASK=w60,dE=1, dQ=0.1,Qmax=2.0,Ib=w4,w_el=w9) & w22=brisp_norm(w22) 
   Calculate S(Q,w) (the ki/kf correction is taken into account) and normalise to the monitor counts. 
   dE=E step (default: 0.5 meV) - dQ= Q step (default: 0.1 Å
   -1) - Qmax (optional) : to limit the calculation to a given Q. 
   Ib=w4 contain the ambient background as determined in 6). 
   w_el=w9 (optional) : the evaluation of the elastic channel is made for each of the 160 x 128 pixels, and can be done either using the elastic signal from the sample in w2 (parameter w_el not needed) or, alternatively. using that measured with another sample (e.g. vanadium in w9). 
8. w32=total(w22,1) 
   Energy integration of S(Q,w) to get S(Q) 
9. 9) out_Sqw, w22, file='namefile' 
   Export S(Q,w) in w22 in ASCII "Q files" (one file for each Q cuts) 

These and other more specific BRISP macros (written since 2005) are part of the ILL lamp package that  can be freely downloaded here.

For any  inquiry: formisano(at)ill.eu