Question about acoustics -- Wine glasses filled with varying amounts of water

[Sylvester McBean]

Here’s my experiment: I have a wine glass half full of water. I tap the edge with a metallic object to make a ringing sound. I recently bought a nice microphone and some acoustic software (Spectrum Analyzer Pro Live) and I’m trying to analyze the resulting sound to determine the resonance frequency. My hypothesis was that the resonance frequency would change in a predictable way (albeit not linear, given the shape of the glass) depending on the amount of water in the glass. I’ve linked to a short video of me doing this. I was surprised to find that regardless of the amount of water, the peak was usually the same 4-5 frequencies. There did seem to be a relationship between how hard I tapped the glass and which frequency peaked.

So I guess my question is this: Assuming I was able to standardize the impact on the glass and tap it in exactly the same place with exactly the same velocity each time, what changes would you expect in the resulting sound with respect to the amount of water in the glass? The pitch clearly changes depending on the water, which is why I was surprised to see the same peak frequency numbers. I guess maybe I don’t fully understand the relationship between pitch and frequency. Would you expect the amplitude to change? Logically it seems like the more water in the glass, the less the glass can vibrate and thus the lower the amplitude. Are there any other characteristics that would change with the depth of water? My goal is to come up with a formula that can approximate the water in the glass from the sound alone.

https://www.dropbox.com/s/xt8ij2gs8mdesan/resonance.mp4?dl=0

 

[anorlunda]

the same 4-5 frequencies

What does that mean?

 

[Sylvester McBean]

What does that mean?

I mean the same several numbers would come up again and again. If you watch that video, you'll see what I mean. For example, when I first started the experiment the glass was half full and the most common frequency I saw was 301.13. Then I added more water to the glass and I expected to never see 301.13 again, but that wasn't the case. It was a little less frequent, but 301.13 was still among the common frequencies I would see.
I hope that makes sense...

 

[Mister T]

Here’s my experiment: I have a wine glass half full of water. I tap the edge with a metallic object to make a ringing sound. I recently bought a nice microphone and some acoustic software (Spectrum Analyzer Pro Live) and I’m trying to analyze the resulting sound to determine the resonance frequency.

Before you do that ...

Assuming I was able to standardize the impact on the glass and tap it in exactly the same place with exactly the same velocity each time, what changes would you expect in the resulting sound with respect to the amount of water in the glass?

Listen to the sounds created when you do this and make your own guesses as to what the measuring instruments might reveal.

 

[Sylvester McBean]

That's the thing. Subjectively, it's not hard to tell the difference between a full and nearly empty wine glass. What I'm looking for is objective data that will back this up.

 

[Mister T]

That's the thing. Subjectively, it's not hard to tell the difference between a full and nearly empty wine glass.

Can you describe that difference?

What I'm looking for is objective data that will back this up.

If you want to "back up" the claim that the sounds are different, display graphs of the two and point out the differences. That's objective data that supports your claim.

On the other hand, if there's something you hear that you can describe, perhaps that description can lead to a more meaningful analysis of the data.

 

[Sylvester McBean]

FYI, this isn't a homework question. ;-) I'm a grown man asking because it relates to a project I'm working on.

I understand that my question was a bit nebulous, so I'll distill it down. Let's assume the walls of the wine glass were perfectly cylindrical, and let's assume that the glass was struck in exactly the same way each time. I would expect that the amplitude of the vibration and the duration of the vibration would display a linear relationship to the amount of water in the glass (i.e., more water, lower amplitude and shorter duration). By 'duration' I mean picking some arbitrary amplitude level and measuring the amount of time between when the glass is struck and when the amplitude drops below that particular level. Another way to think about the duration aspect is to imagine a vibration sensor with a set threshold for firing. The duration would be the amount of time between when the glass was struck and when the vibration sensor stops firing. I would expect a linear, inverse relationship between this time value and the amount of water in the glass.

Am I thinking about this right??

Thanks for the replies, I really appreciate it.

 

[Tom.G]

Just one of many:


found with: https://www.google.com/search?&q=play+a+tune+on+water+glasses

 

[Sylvester McBean]

Yes Tom, I’m aware that the tone is different depending on how much water is in the glass. That doesn’t answer my question, which again is:

Are amplitude and vibration duration inversely correlated with water in the glass, such that you could calculate water in the glass from the sound alone?

 

[Toby Marshall]

You are not taking into account the fact that the glass itself has fixed resonant frequencies, I think. This is a quite different experiment than blowing across the top of a bottle filled with different amounts of water, which varies the amount of air in the enclosed space, but it which there are no idiophonic resonant frequencies due to wall vibrations. The glass itself will have a number of impedance peaks (at which the glass might prefer to vibrate). As you fill the glass, the water will damp certain of these impedance peaks, forcing the glass to vibrate at the next primary peak in its spectrum, but other strong peaks will remain. Because the glass is open, at the top (an exit diameter and port length are important to Helmholtz resonators) you will not be getting any strong resonances from the air inside the glass (as you do blowing across the top of a bottle acting as a Helmholtz resonator.)

IOW your glass is not a non-linear generator, and prefers to vibrate at certian frequencies and not others, and the water can affect which of those are damped, but there will still be strong peaks that will probably be present no matter how much water is in the glass. I'll bet if you change glasses, you will get different fixed peaks. Given this, you are going to have a very hard time eliminating the effects of the fixed impedances of the glass itself in trying to formulate a way to tell how much water is in the glass from the sounding tone alone.

 

[Sylvester McBean]

Thank you for the excellent reply. What if we changed the experiment to an enclosed space, like an oil drum or something. Assume that I'm banging on the side wall, at the very top. In this situation would the resonance frequency change depending on how full the drum was?

 

[anorlunda]

Thank you for the excellent reply. What if we changed the experiment to an enclosed space, like an oil drum or something. Assume that I'm banging on the side wall, at the very top. In this situation would the resonance frequency change depending on how full the drum was?

Maybe, but not in any simple way. It would depend on whether the drum has vertical and/or horizontal ribs, on the shape, on the thickness, on the lid and so on.

It sounds like you are confusing air vibrations that vary with volume (as in a trombone, or blowing across the mouth of a bottle) with vibrations of the enclosure itself.

Have you actually tried the wine glass experiment?

 

[Mister T]

The duration would be the amount of time between when the glass was struck and when the vibration sensor stops firing. I would expect a linear, inverse relationship between this time value and the amount of water in the glass.

If a relationship is linear it can't be an inverse relationship. And vice-versa.

Am I thinking about this right??

Apart from the issue I raise above, no. You don't mention any thinking at all. You just state a conclusion. I have no idea why you'd think there would be any relationship at all. And you say it's struck in the same place every time, but the relationship may depend on which place is chosen.

Perhaps if we knew why you want to know this we might be better able to answer your question.

 

[Sylvester McBean]

I state my hypothesis, and I give reasons for my hypothesis (increased pressure from the water on the walls of the vessel will dampen vibrations, and thus decrease the amplitude and duration of a vibration from an external source). This dampening force, if not linear, will at least have a predictable impact on the external vibration, such that the level of water could be estimated from measuring the sound/vibration alone.
I'm having trouble getting an opinion on this statement.

 

[anorlunda]

@Sylvester McBean , it sounds like you are are trying to make a liquid level sensor that works by thumping the container.

Rather than ask a theoretical question, why not try an experiment? Take a container and thump it. Measure the frequencies and amplitudes of the response. Add some liquid and repeat. If you find a relationship that varies continuously and predictably (but not necessarily linearly) from empty to full, then you could use analysis of the sound as a level sensor.

 

[spareine]

Here’s my experiment: I have a wine glass half full of water. I tap the edge with a metallic object to make a ringing sound. I recently bought a nice microphone and some acoustic software (Spectrum Analyzer Pro Live) and I’m trying to analyze the resulting sound to determine the resonance frequency. My hypothesis was that the resonance frequency would change in a predictable way (albeit not linear, given the shape of the glass) depending on the amount of water in the glass. I’ve linked to a short video of me doing this. I was surprised to find that regardless of the amount of water, the peak was usually the same 4-5 frequencies. https://www.dropbox.com/s/xt8ij2gs8mdesan/resonance.mp4?dl=0

In the video, your spectral analysis software doesn't seem to have produced a useful spectrum. An example of what a better spectrum of a wineglass looks like is this: https://www.physicsforums.com/threads/wineglass-singing-vs-pinging.820647/. It was obtained effortlessly using a free app on a smartphone. The measured frequency depends on the height of the water like this: https://www.physicsforums.com/threads/question-about-helmholtz-resonance.955825/#post-6060753

 

[Mister T]

I state my hypothesis, and I give reasons for my hypothesis (increased pressure from the water on the walls of the vessel will dampen vibrations, and thus decrease the amplitude and duration of a vibration from an external source).

What you're describing is a decay in the amplitude.

This dampening force, if not linear, will at least have a predictable impact on the external vibration, such that the level of water could be estimated from measuring the sound/vibration alone.

Okay.

I'm having trouble getting an opinion on this statement.

You don't need an opinion. You need an experiment.

 

[zanick]

Damping (not "dampening" not a word ) doesn't change the frequency, only the amplitudes and the resonance ring out time. the frequency is related to the natural frequency of the glass of water WITH the water, not the glass alone. water in the glass is analogous to a capo on a guitar which changes the spring mas system. the natural frequency of the system changes. pitch is the same as frequency. you might be getting changes in frequency or no changes in frequency at peak amplitudes (reaction to hitting the glass) because the water is not directly coupled... if you poured molten lead in the glass and did the test again, the results would be slightly different,