Troubleshooting matdyn.x output with imagimary component at Gamma

[ecraddoc]

I am currently working on calculating all the phonon modes at Gamma in a series of materials. My process includes first using the ph.x code to calculate all the phonon modes in all K-space in the primitive cell, then use using matdyn.x, zone fold into a 2x2x2 supercell. The issue I am running into, is that when I calculate the eigenvectors for the primitive cell, all the imaginary components of the optical phonon modes are zero. However, this is not the case for the 2x2x2 supercell eigenvectors. For the next step in my calculations I need this set of eigenvectors, but I am unsure why there are non-zero values in the imaginary component. If anyone has run into a similar issue, or has insights into troubleshooting, please let me know!!!

Here is what I have tried:
Altering the second K-point in the matdyn.x code, or just using Gamma (still, imaginary components are non-zero)
2x2x1, 2x1x2, 1x2x2 cells have mostly non-zero imaginary components, but some some are still non-zero
3x3x3, still same issue.