[FEA] Rebuild the CRE gather stacking with a 2 times interpolated data cube

[Rodolfo Dirack]

🎉 Feature request

Is your feature request related to a problem? Please describe.

The CRE Gather sections generated with the CRE traveltime curve approximation have to much numerical error, and I suspect is because of the numerical precision of the CRS parameters determination or in the determination of the CRE gather traces after the PEF interpolation.

Describe the solution you'd like

• Do the interpolation process twice, in order to improve the CMP sampling in the seismic data cube, to get the CRE gather traces precisely.

• Use variables of type double instead of float.

• Use a local optimization method to improve vfsa as referenced in the issue #8.

Describe alternatives you've considered

I've considered to use the CDS traveltime approximation. Though, I have to explain why does it work as a substitution of the CRE gather traveltime equation.

Additional context

CDS traveltime approximation is when you use the CDS condition (RN=RNIP) in the a CRS traveltime approximation, and it's usually used as a diffraction traveltime approximation. I will have to explain why this diffraction approximation can be used (and it is better) to stack reflection data.