What is that *thunk* I hear? Can anything be done about it?

[funkapus]

Originally Posted by N+1

I'm 70. I know I'm losing some of the high frequencies, but I don't think that's a reason for hearing exaggerated clunks. Or is it?

I'm 53. As your image from Audacity shows, age is unlikely to be the reason why people hear it; there's clearly something there. But it could play into why some people find it easier to hear than others. I don't know the age distribution of people who hear it vs. people who do not; but I wouldn't be surprised if you and I find it harder to hear some of the higher-frequency overtones that would otherwise contribute to the overall amplitude, and thus would perhaps mask the unpleasant thunk a bit.

[N+1]

Originally Posted by funkapus

One other thing I was thinking about doing in Audacity that you might do if you have the time: highlight the time period right around the thunk and ask Audacity for a power spectrum (frequency analysis). Do the same thing for the previous tones on the lower strings. One expects to see the fundamental, the overtone series at lower amplitudes, and a little bit mixed in from the strings plucked previously that are still vibrating (from their sustain). I'm curious whether the power spectrum for the open G string looks significantly different from the others (e.g. a *much* stronger fundamental, or an unusual array of overtone amplitudes, or whatever)

OK, I've obtained spectra for the clunking string (top image) and the low E string (lower image). She's playing an E chord, so the two notes will be G# on the G string (207.7 Hz) and E on the E string (82.4 Hz). We can see those fundamentals in each plot. On the clunking string the fundamental is easily the highest peak in the spectrum. But on the E string the amplitude of the fundamental is lower than the 2nd and 3rd harmonics, and I wonder if this is evidence in favour of Charles's theory about buzzing lower strings - that is, fret buzz is limiting the size of oscillations of the fundamental on those lower strings? In other words, the lower three strings are not vibrating as they should, and the G string is ringing out loud and clear....

[N+1]

Just an extra thought: the so-called thunk is not the open G. It's a fretted G#. So whatever else we might think, this cannot be a nut problem!

[justonwo]

Originally Posted by N+1

OK, I've obtained spectra for the clunking string (top image) and the low E string (lower image). She's playing an E chord, so the two notes will be G# on the G string (207.7 Hz) and E on the E string (82.4 Hz). We can see those fundamentals in each plot. On the clunking string the fundamental is easily the highest peak in the spectrum. But on the E string the fundamental is lower than the 1st and 2nd harmonics, and I wonder if this is evidence in favour of Charles's theory about buzzing lower strings - that is, fret buzz is limiting the size of oscillations of the fundamental on those lower strings? In other words, the lower three strings are not vibrating as they should, and the G string is ringing out loud and clear....

That's pretty cool!

[funkapus]

Originally Posted by N+1

OK, I've obtained spectra for the clunking string (top image) and the low E string (lower image).
(snip)

Thanks for doing this. Quick question: how did you capture the sound from the video?

[N+1]

Originally Posted by funkapus

Thanks for doing this. Quick question: how did you capture the sound from the video?

Set the video paused just before the section you want
Open Audacity
Hit Audacity's red record button. It begins to record.
Start the video
Stop the video
Stop Audacity and edit the recorded track as required.

Full of fun and frolics.

[funkapus]

Originally Posted by N+1

Set the video paused just before the section you want
Open Audacity
Hit Audacity's red record button. It begins to record.
Start the video
Stop the video
Stop Audacity and edit the recorded track as required.

Full of fun and frolics.

OK. So it passed through the sound drivers of whatever OS you use.

I was contemplating pulling the audio track of the video(s) directly from youtube into whatever best format YT will give it to me in.

[N+1]

Originally Posted by funkapus

OK. So it passed through the sound drivers of whatever OS you use.

I was contemplating pulling the audio track of the video(s) directly from youtube into whatever best format YT will give it to me in.

That would get you closer to the original audio, though I'd be surprised if it gave a very significantly different result in terms of what we're listening for. I can't hear a difference between my extracted audio recording and the video itself.

The most interesting aspect, for me, is that the thunky G# has a spectrum that broadly looks alright, whereas the low E doesn't look right at all. And now, on re-listening to the video, I'm wondering about what my ears are telling me, and remembering an old speaker design trick, of lifting the upper bass slightly to 'fool' the brain into thinking it hears lower bass notes than are actually present. In this case I 'feel' as if I hear that fundamental 82 Hz E string, even though the spectrum tells me I'm hearing more of the two higher harmonics.

If Charles Tauber is right, then the G# is the only note that's ringing correctly (as it broadly seems to be, judging from the spectrum), and the three lower strings are all messed up to some degree. And what we hear as a 'thunk' would be merely the consequence of hearing one note louder (and purer) than all the others (as someone pointed out above).

[frankhond]

Originally Posted by funkapus

OK. So it passed through the sound drivers of whatever OS you use.

I was contemplating pulling the audio track of the video(s) directly from youtube into whatever best format YT will give it to me in.

The *drivers* are irrelevant unless you use EQ somewhere, they just pass the audio to the AD converters on your soundcard. Also recall that youtube audio is compressed with some lossy algorithm. You already have a compromised frequency spectrum. A frequency plot can show you the rough profile of the sampled fragment, I wouldn't expect more insight from bypassing the drivers.

[frankhond]

What I hear is that the fretted G# has a strong resonance compared to the other notes.

It could be the room. I have an issue where certain open strings (D and G) sometimes have an irritating ringing quality when playing at home. This goes away when bringing the guitar to the shop or a friends house.

It could also be a wolf tone. Each guitar has inherent resonances, a luthier tries to minimize these by tuning them to inbetween notes that are not often played (I read that slightly sharp of F# is a popular choice). Still this doesn't eliminate the problem entirely, and when playing these notes often appear as "thunks" to me as a player. So the phenomenon is real.

However, no listener has ever complained that my guitar is making thunks. When you are playing music, that stuff goes away. Part of that is technique, when I know there is a thunk note coming up, I just compensate. But these things are also part of the guitar as an instrument.

If you are jonesing for a guitar but it has a thunk that is unbearable to you, don't buy it. It could well be that this particular lovely 1931 OM is in the shop because the owner couldn't stand the thunk.

Once Mary starts playing a song the thunk becomes irrelevant, I listen instead to the player's dynamics and phrasing. One can notice that the guitar becomes very resonant in the lower mids (what many people call "bass") around frets 5-7. Also there is a hard edge to the recording that emphasizes string noise and buzz, probably caused by an untreated room with hard walls. Clearly the recording setup is suboptimal, I wouldn't evaluate this guitar based on the sound in this video. That goes for most youtube material.

[N+1]

Originally Posted by frankhond

Once Mary starts playing a song the thunk becomes irrelevant,

I don't find this. When she's playing, the thunks completely destroy the musical qualities of the performance, for me. They continually punctuate the music like sledgehammer blows would, and are just as unignorable.

One thing I think needs considering is that the audio spectrum for the bottom E is not what I'd expected. Why is the fundamental so diminished in amplitude?

[MHC]

Thought I'd add this to the mix.

Same guitar and somewhat same recording conditions (same room but we can't know about mic placement) -- different player.

I wonder if the "thunk" is noticeable on this one for those who hear the "thunk."

[N+1]

I wonder if the "thunk" is noticeable on this one for those who hear the "thunk."

I hear quite a few disconcerting thunks (and mini-thunks) on this - not so prominently perhaps, but I hear enough to make me not want to listen. There are two substantial ones at 0.56s and 0.59s for instance.

[frankhond]

Originally Posted by N+1

I don't find this. When she's playing, the thunks completely destroy the musical qualities of the performance, for me. They continually punctuate the music like sledgehammer blows would, and are just as unignorable.

One thing I think needs considering is that the audio spectrum for the bottom E is not what I'd expected. Why is the fundamental so diminished in amplitude?

Where are they? The thunk in the beginning is fretted G# near the nut. Most of what she plays later is up and down the neck. I do hear certain notes popping out. This is a standard feature of the blues/ragtime playing style.

[charles Tauber]

Originally Posted by N+1

I hear quite a few disconcerting thunks (and mini-thunks) on this - not so prominently perhaps, but I hear enough to make me not want to listen. There are two substantial ones at 0.56s and 0.59s for instance.

I hear a mass of buzzing and percussive striking. I chalk it up to the "style" of playing and technique (and, perhaps, to the setup of the guitar). Both players are demonstrating a similar genre of music on that guitar. I appreciate the style, but it isn't one that particularly appeals to me.

I have no doubt at all that the instrument would sound very different if played by other people with a different style of music. People don't want to hear/believe that so much of what an instrument sounds like has to do with the person playing it and how they play (aka technique).

Based on the two recordings, it isn't likely an instrument I'd want to own. It doesn't have the sound characteristics that I look for. I'm not swayed by the fact that it is a "vintage Martin". For others, and what they look for, it is their dream guitar.