Unless you are writing your own FFT, you probably can't easily use QVector<double> in a routine that expects double pairs. However, if all you need is a pointer to a C++-style array to pass to a library function, then use the QVector::data() method to obtain a double * pointer.

Most FFT libraries I have seen expect a single double * array of amplitude values and have an implied frequency that starts at zero. (The frequency array is in fact irrelevant in this case - the FFT expects the amplitude array to represent values at equal frequency intervals).

But maybe I don't understand your question.