Re: [GlueX] Hall-B beamtest

[Rafael Hakobyan <hakobyar@uregina.ca>]

Hi George,

Thanks for the feedback. I agree with you that if I don't change the 
Z-coordinate of the hit position, then I could just
look  at one end only and extract the same conclusion. But since the 
Z-coordinate (entry point) is changing from -9.35mm to 9.35mm because of 
the uniformly distributed  spot size, and plus, I am going to take into 
account the fact that in each readout cell the energy weighted 
z-position can vary in between the Moliere radius size, which you know 
is about 3cm-3.5cm, so we do need to look at both ends.
The geometrical mean, that Elton has suggested, looks indeed much more 
relevant for this study. I'll do that and let's see the outcome.
Thanks,
Rafael 

  
George J. Lolos wrote:

> Hi Rafael:
>
> The geometric mean is indeed a standard way of first level elimination 
> of attenuation length from a long counter when two ends are read out.  
> It also reduces the solid angle effect for such counters, even though 
> the latter is not a factor in the BCAL due to the SciFi nature of the 
> active detector.  Having said this, and as Elton points out, for the 
> specific study you are doing such treatment is not required.  Since 
> you don't change the z-coordinate of the hit position but you only 
> investigate the y-coordinate effects, the symmetry between the two 
> ends - where the energy (pulse height) is read out - makes the 
> arithmetic and geometric means non-factors.  You could just look at 
> one end only and extract the same conclusions. However, once you 
> change the z-coordinate, Elton's suggestion is the only way to proceed 
> in extracting the total energy collected in the BCAL.
>
> George
>
> Rafael Hakobyan wrote:
>
>> Hi Elton,
>>
>> You know I am pondering upon what you have written below and your 
>> explanation
>> sounds to me very reasonable and I'll obtain the distribution of the 
>> geometric mean.
>> It is interesting to compare it with the arithmetic mean that we 
>> already have.
>> Thanks a lot, and I'll let you know as soon as I have results (it 
>> will be very soon).
>> Cheers,
>> Rafael
>>
>> Elton Smith wrote:
>>
>>> Hi Rafael,
>>>
>>> Thanks for producing the plots on energy resolution. For the purpose of
>>> that study the following does not make any difference, but you should
>>> consider using E = sqrt(Ea*Eb) as the measure of the energy instead 
>>> of the
>>> average of exponentially corrected energies. The geometric mean removes
>>> the energy and the attenuation length dependence to the energy 
>>> measure so
>>> it makes it a very robust quantity. If you use other form, you 
>>> introduce
>>> both a dependence on both the position and the attenuation length.
>>>
>>> Cheers, Elton.
>>>
>>>
>>>
>>>
>>> Elton Smith
>>> Jefferson Lab
>>> elton@jlab.org
>>> (757) 269-7625
>>>
>>>   
>>
>
>