jforecast and hard tunes

[Armen Mirzoyan]

Good time of the day,

I am working with the jforecast function for conditional forecasting, I was wondering how exactly the process of hard tuning works. I supply the hard tunes via a separate db, like this: 

f = jforecast(m,db,fcastrange,dbtunes,'deviation',false,'anticipate',false,'plan=',simplan);

It seems that for some occasions, the simplan and hard tunes 'do not go well together.' If we rule out singularity in tuning (simplan vs hard tunes), what could be behind, for example, the following: depending on the simplan, in some cases I am able to replicate exactly the hard tunes in the f.mean, while if I change the simplan, or even the quarter over which the hard tune is applied, I am not able to replicate it in the outcome database. Do you guys have any similar experience with jforecast that could share with me? 

Thanks a lot,

Armen