wish help you to fix your issue Normalise over a a longer scale. You'll need something like an envelope follower with a long release time. If you search for 'compressor' source code, or automatic gain control something will definitely turn up.
fixed the issue. Will look into that further Following examples on other Stackoverflow posts related to word frequency analysis in Python, my program is returning letter frequency analysis and not actually the word.
Java .wav file frequency analysis - incorrect frequency
hop of those help? You don't say how you interpret the results in Excel that you generate with the code above. However, a likely error is misunderstanding the output of the FFTfft.realForward() - which is an array of complex numbers, whose real and imaginary parts occupy consecutive elements, as documented here. If you simply use the index of the array in which the peak occurs, your result would be off by a factor of two. Note that this FFT implementation only calculates up to the Nyqvist rate (beyond this merely yields an 'alias'). Other things to note:
it should still fix some issue A N-point FFT of a signal with a sample rate of 44100 produces frequency bins with center frequencies spaced 44100/N apart from 0 Hz to 44100 Hz. From 0 to the Nyquist frequency of 22050 Hz, there are N/2+1 points inclusive. So if you want the center frequencies then compute i*44100/N where i=0,1,...,N/2.
Setting pitch value when transitioning from FMOD Designer to FMOD Studio
Any of those help In the FMOD QA site there is a formula pitch = 2 ^ ( semitone / 12.0f) which seems to do the conversion I need. At least the end result sounds more or less the same. Of course I would gladly hear from a FMOD expert is this really correct or not.
Frequency Analysis in Python -Print letters with frequency rather than numbers with frequency