geko

I am not frequently considered with Max and hope to find a solution for a problem from which depends quite a lot for me.

a string is recorded by an electromagentic pickup. I try to estimate the frequency with the highest amplitude (in this case not necessarily the f0) on the basis of fffb~. the frequency range is from 60 - 590 Hz. So I run one fffb~ with equally distanced frequencies and a high Q factor. the signals from fffb~ are going through average~ in absolute mode and with a time intervall of 20 (ms) (this should be enough even for a 60 Hz wave). Then there is a list with all averaged signal values and the highest one is searched out by index. But the values are changing after a while to very different indices even if the spectrum has a clear dominating freq around 240 Hz. I need a clear and distinct result. How to achieve this?

As attachment there are my patch and a short soundfile (48 KHz).

Hi,
I am not frequently considered with Max and hope to find a solution for a problem from which depends quite a lot for me.

a string is recorded by an electromagentic pickup. I try to estimate the frequency with the highest amplitude (in this case not necessarily the f0) on the basis of fffb~. the frequency range is from 60 - 590 Hz. So I run one fffb~ with equally distanced frequencies and a high Q factor. the signals from fffb~ are going through average~ in absolute mode and with a time intervall of 20 (ms) (this should be enough even for a 60 Hz wave). Then there is a list with all averaged signal values and the highest one is searched out by index. But the values are changing after a while to very different indices even if the spectrum has a clear dominating freq around 240 Hz. I need a clear and distinct result. How to achieve this? 

frequency detection by filterbank