1) in the case of propagation to the point of closest approach to a space point now we set: vwi = point (vwi = space point itself). Previously the vwi was not filled and so when calculating one of the two axes to define the virtual detector plane with:

// unitary vector along distance

// vpf on track, vwi on wire

TVector3 fromwiretoextr = vpf - vwi;

fromwiretoextr.SetMag(1.);

the result was not correct. Now the first axis is the line where both the space point and the point of closest approach to it lie.

2) In the case when only two steps are performed during the PCA finding the procedure used to give a wrong result, because the FindPCA didn' t take this case into account properly. Now this bug has been fixed and the function works fine.

In case of problems using this new CbmGeanePro please let us know!

Regards,

Lia and Susanna.

]]>

Do I understand it correctly, that you choose the coordinate system such, that the line connecting the pca and the hit coincides with the x-axis of the plane?

Cheers! Sebastian.]]>

I changed this because I saw that in the Propagate(CbmTrackParP *TStart, CbmTrackParP *TEnd, Int_t PDG) you added the "support for poca with points", i.e.:

if(fPCA==1){// point

TVector3 mom(TStart->GetPx(),TStart->GetPy(),TStart->GetPz());

wiredirection=mom.Cross(fromwiretoextr);

}

so I thought you wanted to make your plane perpendicular to the track direction, is this correct?

Previously the vwi was set to (0,0,0) in the case of pca = 1 and so the fromwiretoextr axis (x-axis of the plane) was from (0,0,0) to the pca...

I hope this is ok...

Ciao,

Lia.]]>