Hi Ricardo,
The erroneous warning message formatting means that you are using a version of MRtrix3 prior to 3.0.0, since it was resolved here.
The issue has been discussed multiple times on the forum; see e.g. this thread. It’s not something to be overly concerned about, but you can check the locations of such errors using the 5ttcheck command and quantify the sum of tissue fractions using mrmath sum -axis 3 to see the magnitude of the problem; if the errors are ~ 0.1 or smaller then tracking shouldn’t be affected. This would also likely not appear at all if using an up-to-date installation (or if it does appear it’s a handful of voxels at most).
Cheers
Rob