Finding Fundamental Resonant Frequency of an Instrument Audio Signal

Thread Starter

mwarre25

Joined Mar 22, 2011
13
Hey Everybody,

I am currently working on a basic signals and systems filtering project(attached) and I wanted to ask you guys some general questions as I go along through the project to make sure I'm learning this stuff correctly.

My first question is about estimating the natural or fundamental frequency of a signal graphically.

The wave is that of a bugle. After some research I read that an instrument has a few fundamental frequencies and those frequencies are easily excited. I assume that the high magnitude frequencies on my frequency vs. magnitude graph are multiples of my wave's fundamental frequency.

Is this assumption correct?

With that assumption I used the graphical zoom function to zoom in on all of the peaks and I noticed that they were almost evenly distributed. I assumed that the difference in these peaks was the fundamental frequency.

Is this an accurate way of doing this?

Also, After this step we were required to design an ideal filter to filter out extraneous portions of the wave. This means low frequencies that would sound like hums and possibly high frequencies like cracking right?

Thanks,
 

Attachments

bluemarvin

Joined Apr 19, 2011
8
You cannot tell from the frequency plot alone if it was produced from sequential or simultaneous time signals. But, if the attached time plot is the input that produced the spectrum, it looks like somebody tooted that horn 4 times about 2 seconds per toot which probably corresponds with the 4 tallest spectral groups. From your spectrum those peaks look to be at about 400, 500, 600, and 800 hz...the short peak just below 10,000 rad/sec is probably an overtone or harmonic of the lowest-frequency group. You might correlate those numbers with the notes a bugle can play, such as 'A' being 440 hz. The fundamental frequency of each group is the highest peak...there are 4 different notes.

Regarding the filter, yes you are right that it would be as narrow as possible centered on the fundamental of each spectrum group...if the filter's passband allows any of the adjacent spectral group in, it's not very ideal. The more ideal it is, the harder it is to make though.
 

Thread Starter

mwarre25

Joined Mar 22, 2011
13
Hey bluemarvin, thanks for the response. The description (attached) wants me to find it by just examining the magnitude of the frequency spectrum.

Are you saying that the bugle has more that one fundamental frequency? I thought that it could only produce different harmonics of the fundamental (with overtones).

I have estimated the fundamental frequency to be 123.327321857 Hertz using the method attached.

Thanks for the response as well designnut.

My next task is to design an ideal filter to filter out extraneous portions of the signal.

Regarding bluemarvin's post that "it would be as narrow as possible centered on the fundamental of each spectrum group...if the filter's passband allows any of the adjacent spectral group in, it's not very ideal. The more ideal it is, the harder it is to make though," does, "fundamental of each spectrum" mean like the fundamental multiplied by the desired harmonic? I'm kind of confused at what spectral groups and spectral lines mean...ill keep reading
 

Attachments

Thread Starter

mwarre25

Joined Mar 22, 2011
13
Hey,

Yea, I could have asked him but I wasn't really able to get to his office. I finished the project though and I'm building the filters. I'll post pics later. I really enjoyed this project.
 
Top