In the book Automatic Control Systems, Benjamin C.Kuo, 7th edition, on page 548, he says: https://imagizer.imageshack.us/v2/720x154q90/540/ML8zmu.jpg [Broken] He is doing an analysis of the following transfer function: http://imagizer.imageshack.us/a/img911/4839/XPKxoO.gif [Broken] Mr is the maximum value that | H (jw) | can reach with w ranging from 0 to infinity. He says that if ζ is negative, the system is unstable and the value of Mr ceases to have any meaning. I disagree with that. Whereas if ζ is negative, it is clear that this puts the two complex poles to the right side of the real axis, but the function | H (jw) | is exactly the same in the cases of ζ be positive or negative and it is not by fact a transfer function having its poles right that Mr ceases to have meaning. Mr will be: http://imageshack.com/a/img538/9372/tFTOaZ.gif [Broken] (as discussed on page 546) independent of ζ to be negative or positive. Ie, it is not because the system is unstable to a step in the time that their analysis in the frequency domain loses meaning. I would like to be corrected if my view is wrong. Thank you.