shear strain

Dear All

Looking for a methodology to convert shear stress-shear strain data from a torsion test in large deformations into true strain-true stress curve, I found two schools of thought: 

and

I found also in recent papers a strong debate on which should be the correct method. For example: