Hi Kirk,

here are some polished figures:
PhotonNumberHelicity19b - the undulator spetrum relabeled...
pol2122_dark - the polarization plot relabeled
central_csi_160 - the energy deposition in better quality(3 times with 
different x axis title...)
asymmetries_140_2_b.eps - the single asymmetries relabeled


Concerning the analysis:
Eq. 24 describes the background subtraction in general, were n_s,b are 
the number of undulator-on/off events or beam pulses or triggers. In a 
"normal" cycle this number is after cuts etc. in the order of <1500.
Technically, the background subtraction and the normalization is done 
(has to be done - we have to combine the correct signal-background pairs 
from the CsI and the silicon counter) in one go according to eq. 31-33 
in the Appendix A.5.
We found out, that it is equivalent
- to normalize to the photon flux and correct with the asymmetry 
measured in P1
- or directly normalize to P1
The number of positrons does not enter here - we are using the 
un-calibrated readings from the P1 counter for the normalization/correction.
I hope, this explanation helps.
Up to which detail do we want to describe the analysis, do we want to 
come up with formulas...?
Best regards
Ralph

p.s. I wish everybody Merry Christmas and a happy new Year...!



Kirk McDonald wrote:
> Since we have some room left in the PRL draft, it might be good to 
> include a bit more about how the positron asymmetry analysis was done.
>
> Looking at the long paper around pp. 75-76 I can't quite reconstruct 
> what is being said there.
>
> I'm looking at eq. (24) [which to me is an "interesting" technique 
> that may deserve mention in the PRL].
>
> Does s_b = energy observed in a CsI crystal during an undulator-on 
> event (where this energy is "calibrated" into MeV units, but NOT yet 
> "normalized" to counts in P1)?
>
> It sounds like calculation (24) is performed prior to using any 
> information from counter P1.
>
> Then (maybe) the calculated energy E+ or E- is somehow "normalized" 
> with counter P1.
>
> But it's not clear what this means.  Or why we might want to do this.
>
> ???   I gather that in eq. (24) n_s ~ n_b ~ 1500.
> The number of positrons in P1 per pulse was about N_s ~4e4 for 
> undulator on. I can't find a statement in the long paper as to the 
> number N_B of "positrons" in P1 during undulator-off pulses.
>
> Anyway, I gather that we used the numbers
> M_S = sum_{i=1,n_s} N_S,i ~ 6e7
> M_B = sum_{i=1,n_b} N_B,i ~ ????
> to modify equation (24) in some unspecified manner.
>
> Clearly, I need a "reminder" as to what's going on here -- which hint 
> we might want to pass on to readers of the PRL....
>
> --Kirk
>
>
>


-- 
--------------------------------------------
Ralph Dollan	Ralph.Dollan@desy.de

Tel:  +49-33762-7-7508
Fax:  +49-33762-7-7330
DESY
Platanenallee 6
15738 Zeuthen
--------------------------------------------