Modify SolidAngle to take into account object 'shadowing' for PSD's

[Stuart Campbell]

At the moment, the SolidAngle algorithm doesn't return the correct value if part of the detector is obscured by another component. This is most common in a detector with multiple elements, i.e. PSD tube, where one pixel obscures the bottom/top of the next pixel.

[Stuart Campbell]

(In [9846]) Commented out use of SolidAngle in MedianDetectorTests as it can give false (good) results for pixels. refs #2596

[Stuart Campbell]

(In [9878]) Updated Excitations Reference results to take account for changes to determined mask. refs #2596

[Nick Draper]

Actually there is a simpler approach we can take here that will solve the common problem we have without affecting performance adversely.

This problem only appears significantly for PSDs. Which are short cylinders stacked on top of each other. The problem occurs because of the end-cap of the cylinder becoming visible as the angle increases. For non cylinder detectors this is easily solved by giving the detector a very small depth.

We could code in a special case for cylinders (they already use a special case for rendering). During the object construction it can be recorded that it is a cylidner (already occurs) and that it is part of a larger tube (not done yet), in which case we can use an analytical formula to give the solid angle in this special (but common) case. ​http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B73DN-470W9GH-8P&_user=910841&_coverDate=06/01/1972&_alid=1671689142&_rdoc=5&_fmt=high&_orig=search&_origin=search&_zone=rslt_list_item&_cdi=11475&_docanchor=&view=c&_ct=55817&_acct=C000047841&_version=1&_urlVersion=0&_userid=910841&md5=53e0f1d010525515b77334b22607531a&searchtype=a or

​http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TVT-48DXNGW-2&_user=910841&_coverDate=07/31/2003&_rdoc=1&_fmt=high&_orig=gateway&_origin=gateway&_sort=d&_docanchor=&view=c&_searchStrId=1671689142&_rerunOrigin=google&_acct=C000047841&_version=1&_urlVersion=0&_userid=910841&md5=457c7d30b42a48ae5dae9e1cd0cac4f8&searchtype=a

[Nick Draper]

Bulk move of tickets at the end of iteration 28

[Martyn Gigg]

(In [11674]) Modify the SolidAngle calculation for cylinders to EXCLUDE the end caps so that it closely approximates the effects of shadowing for stacked cylinders. Also cleaned up some indentation in renderinging code and made an error message clearer in IPropertyManager. Re #2596

[Nick Draper]

Accepted and assigned tickets moved at iteration 29 code freeze

[Stuart Campbell]

Works ok.

We should now also start re-using it in out diagnostics (see #3742)

[Stuart Campbell]

This ticket has been transferred to github issue ​3443