Page 1 of 1

Spectrum Analysis of Musical Notes

Posted: Fri Feb 19, 2010 11:45 pm
by England
Hi,

I have searched through pages of the search results but couldn't find a topic on a problem similar to mine.

My goal right now is to record a scale on my bass guitar and to determine the frequency of the notes (fundamental and overtones/harmonics).

My problem is that while I am playing a certain note, in this case : C (should be C1, since I'm playing on the A1 string), I see the spectrum analysis seeing the note as B (oddly too: it's loudest B2, instead of B1). I was just wondering if this had been noted before? (I have not tried another microphone, I need to search for another one if that's the case).

Here is a SS
Image

My second question is that I could not achieve this range of frequencies (from 100 - 1000Hz) unless I changed Edit -> Preferences -> Quality -> Sampling -> Default Sampling Rate. If it was set at the default at ~16000, then I got a huge range of frequencies (completely blurred the important lower frequencies). But it's weird, since the sample rate seemingly has nothing to do with the frequency of the recording itself. (If this question is much longer and complicated than I thought, then maybe I will separate it as a separate post).

Thanks!

England

Re: Spectrum Analysis of Musical Notes

Posted: Fri Feb 19, 2010 11:57 pm
by kozikowski
Like this one...

http://kozco.com/tech/audacity/piano_G1.jpg

You'll need to change some of the settings on the spectrum tool like the ones on the bottom of that illustration. Also, you need to know that the amount of information goes way up as you grab and stretch the spectrum window larger and larger.

You might also need to get a little louder. Your peak note is -30dB and that's seriously quiet. The signal level goes down by half every 6dB. So that's 6, 12, 18......etc.

Koz

Re: Spectrum Analysis of Musical Notes

Posted: Sat Feb 20, 2010 12:39 am
by England
Thanks Koz, I'll give those a try.

What about the playing-C-reading-B? Have you heard of that previously?

England

Re: Spectrum Analysis of Musical Notes

Posted: Sat Feb 20, 2010 12:51 am
by kozikowski
Yes, but not that bad. It's possible to have a sound card with a damaged digitizer or time base managers and produce some really odd results. Damage here sometimes results in not being to perform to yourself by overdubbing or to a standard click or rhythm track.

If you record a note or two, export as WAV and play it back on another computer, do you get the same note? Splitting computers like that is sometimes a good way to figure out where the damage is.

That and if you have a serious fuzz guitar, you can totally get harmonic peaks bigger than the main note. What would be very unusual, however, is if you got that while you were playing a high quality, hollow-body, wooden Gibson.

Koz