In my project I need to detect the dominant frequency going into a mic (I'm using a yamaha keyboard or whistling to test it).
My current implementation essentially copies the FFT minim example, and looks for the band with the highest strength. If I'm using a pure wave of some kind it works fine, but on some instruments settings, the signal indicator jumps between the correct one and a much higher one.
Now I'm assuming this comes from harmonics in the instrument, and not a more general noise issue (though I do need to look at that further). So my question is whether there is a mathematical/programmatic way to remove harmonics from a signal.
Thanks
My current implementation essentially copies the FFT minim example, and looks for the band with the highest strength. If I'm using a pure wave of some kind it works fine, but on some instruments settings, the signal indicator jumps between the correct one and a much higher one.
Now I'm assuming this comes from harmonics in the instrument, and not a more general noise issue (though I do need to look at that further). So my question is whether there is a mathematical/programmatic way to remove harmonics from a signal.
Thanks
1