Assignment-3 Spectral Analysis

1. Flute Analysis

My first analysis was for a sustained flute note complete with vibrato tones. The spectrum has a very clear primary frequency around 1770 htz. There are multiple lesser harmonic frequencies that are also very clear and show a regular downward trend in decible levels.

The spectrogram also shows the primary harmonic frequency very clearly as well as clear subsequent harmonies. The vibrato in the flute noise is also very visible in the vertical dashes. The image resolution was best in a 1024 window for both vibrato and harmonic qualities.

2. Hammer and Concrete Block

The sound I used was the sound of a metal hammer hitting a concrete block. producing a very high pitched tinkling sound. This was reflected in the spectrum with the highest harmonic frequency being around 1660 hz. After that the frequencies are fairly difficult to discern although they show a sloping downward trend. It becomes very difficult to find other significant harmonic frequencies due to the short pitched nature of the sound.

The spectrogram shows the hits as very dramatic and lacking a large amount of resonant reverberations. The main frequency is very evident like in the spectrum but there are not a great deal of other significant frequencies. The short sharp nature of the sounds made isolating a good sample very difficult because there were not many times where there was a great deal of sound.

3. Bean Belt

My third sound was an African instrument made of dried bean pods that is shaken to make a rattling noise. Like the concrete block, the harmonic tones of the bean pods were difficult to calculate. The most relevant frequency was between 1765-1808 hz. Like the block, the beans also had low decibel levels under 500 hz.

In many ways the spectrogram was more difficult to analyze than the concrete blocks. Because the separation between sounds was very fast and not very delineated, the noise from the resonant frequencies blocks out the oscillations in the 1770-1800 hz range. At higher frequencies, these delineations are more visible meaning that if this sound was put through a relevant high-pass filter, the oscillations would be way more audible.

4. Rooster

A rooster call has a very recognizable sound and I chose to use the middle of the call where I felt the signal had the most regularity for my spectral analysis. The main frequency came in at about 1800-1850 hz. This was not a very clear call for a bird and other relative harmonics show up more in groups of similar oscillations rather than specific harmonic notes. This trend is definitely reflected in the spectrogram.

One of the peculiar aspects of this particular recording is the grating noise that the bird makes in the middle of the sound. The spectrogram shows that the middle of the call is much more muddy and has fewer harmonics than the beginning or the end. Since the spectrum readings were taken from this muddy section you can see the harmonics clumped together in sections rather than nicely arrayed like in the flute analysis. The main harmonic frequencies are a group of three in particular that are echoed in the spectrum.

-Razz