I have not worked with NEST-DORIS software, but if the issue affects not only the software, I can say that in the interferometry there are different approaches to the processing of the spatial phase distributions. Many of them mentioned in the introduction of the attached paper.

Maybe it will be useful for you. Probably, the problem caused by phase wrapping?

Article Digital off-axis holographic interferometry with simulated wavefront

The interference phase is a mixing between the object phase and the reference phase

Delta phi = phi_obj –phi_ref

The main task done by the proses is compensating for the reference phase (which assumed to be known during the experiment).  This could be achieved digitally or experimentally (by recording an interferogram without a test object).

So obtained intensity holograms should be processed until one get complex amplitude b (phase and amplitude) using phase shifting techniques (like spatial carrier method or temporal phase shifting). Then the one need to multiplying complex amplitude obtained from interferogram of the object and the reference with the conjugate of complex amplitude of the reference wave (interferogram without the test object or thedigital reference wave).

The reference phase may also be recorded at the end of the procedure. It may also be refreshed during real-time recording of object phase or synthesized. Different other useful mehods are described in Journal of Solid Mechanics and Materials Engineering 01/2008; 2(6):695-706 - and some of them might be adapted.

I hope I've been able to help !

Usually when analysing an interferogram pattern, the result will be tilted due to the presence of carrier frequency.. therefore, subtraction of the reference phase might mean removing the tilt from the result.