Jump to content
Maestronet Forums

Secret Knowledge and violins


Craig Tucker
 Share

Recommended Posts

Don't go cooking varnish on your boat :o

Yeah, I was thinking that, too...there have been some tragic accidents with fire aboard yachts, workboats, houseboats and ships.

Hit the beach for varnish making, OK?

I recall a very nice all-fiberglass (GRP for our UK friends) medium-size commercial vessel that caught fire off the coast of Oregon. The fishermen were very lucky to have survived, and that only because they had a life-raft of some sort... the boat was a total loss.

Link to comment
Share on other sites

  • Replies 630
  • Created
  • Last Reply

Top Posters In This Topic

Don't go cooking varnish on your boat :o

Yeah, I was thinking that, too...there have been some tragic accidents with fire aboard yachts, workboats, houseboats and ships.

Hit the beach for varnish making, OK?

I recall a very nice all-fiberglass (GRP for our UK friends) medium-size commercial vessel that caught fire off the coast of Oregon. The fishermen were very lucky to have survived, and that only because they had a life-raft of some sort... the boat was a total loss.

Fire aboard must be scariest thing at sea, I'll tell you if it wasn't for the violin community warning in multiple posts I could have done it. None of the old varnish books I had read before had explained the dangers as well as all of you, thank you very, very much.

My friend from another boat lends me a portable cooker to cook at the beach.

Once again thank you for the knowledge and for all the learning the past few months.

Link to comment
Share on other sites

I have seen only FFT programs that give one graph (the real part). Are there others that give the imaginary part? The phase angle would be arctan(im/real) wouldn't it? How would you use this information?

SpectraPlus does both. Stoppani's software also uses the phase information, but I think it is plotted in two planes (real and imagianry part) instead of showing only the phase. I have not explored his software so well, at least not beyond the modal analysis part.

The phase between the hammer signal and the accelerometer signal from e.g. behind the bridge foot, and hammered at the bridge side would say something about how the resonances do combine. E.g. the B1- and the A0 sum between them is shown in the phase plot, also that the B1- and B1+ 'phase each other out' between them can be seen there (as well as in the amplitude plots).

E.g. reading the phase relation between the signal from a tap on the back plate and the accelerometer on the top would say something about any breathing or bending qualities of a resonance in a violin. Breathing behaviour (plates moving in the opposite direction) is likely to give a stronger sound radiation than bending (plates moving in the same direction).

Modal analysis can give more details in that process as the information is built up from data from some 200-250 such hammer tap positions all over the plates, recorded at the single accelerometer position. Jansson also used the phase information to determine where the body and bridge hill is in a fiddle admittance spectrum.

Link to comment
Share on other sites

I am a escapee from the recording studios, now living aboard with the boss (the wife - I'm the captain....because she says so) and having lots fun with varnish (I hope t does not turn into an obsession, I'll have nowhere to escape to)

Best to you, Carlo. If you have a lot of experience with recording, and sound production and sound reproduction, you'll probably be a hugely valuable asset here.

Link to comment
Share on other sites

The 'math' behind harmonics is well understood, but not violin geometry. In this case, acoustic violin design IS geometry which effectively IS the "oscillation system". The answer is math/geometry.

I suspect you believe corpus resonance modes are like the random numbers generator rather than a function of instrument design geometry.

Jim

I respectfully disagree. The string is the oscillating device, and the corpus is nothing more than an acoustic bandpass filter and driving device to couple the filtered mechanical waves to the air. There are resonances involved as in any filter network, but dampened and incapable of sustaining oscillation.

If you figure some way of specifically phase tuning a wooden box across 5+ octaves, I sincerely have nothing but the highest level of awe and respect for your knowledge and abilities.

Link to comment
Share on other sites

I respectfully disagree. The string is the oscillating device, and the corpus is nothing more than an acoustic bandpass filter and driving device to couple the filtered mechanical waves to the air. There are resonances involved as in any filter network, but dampened and incapable of sustaining oscillation.

If you figure some way of specifically phase tuning a wooden box across 5+ octaves, I sincerely have nothing but the highest level of awe and respect for your knowledge and abilities.

Bill,

Perhaps one need only look for where the violin creates 90º phase shifts ... as between sine and cosine.

Jim

Link to comment
Share on other sites

That's basically kindergarten stuff, versus what Bill has been talking about.

Bill and probably thousands of others basically look at the strings as thee oscillating device with corpus creating mechanical waves coupled to air. Kinda like a speaker diaphragm, huh? That's one way of looking at it. The researchers also suggest only ~30% of corpus vibrations as seen in 3D animations contribute to the sound radiated and heard. REALLY!?!

The other way of looking at it is to drop all this mechanical wave coupled-to-air nonsense which may simply be analogous to distorted lip-flapping by a helmetless motorcyclist traveling at high speed. It ain't the lip-flapping that gets you from Point A to Point B [it's those spinning wheels!}.:lol: Let strings rightly be an oscillating device, oscillating 'air' which is compressed & rarefied, then projected to ambient air by pressure gradients while maintaining varying velocity and phase relationships. This is what Amati's genius was trying to accomplish.

Jim

Link to comment
Share on other sites

Mechanical wave coupling to air is certainly not nonsense; I don't know of any other method of producing audible sound. I'm sorry, I don't follow your analogy between a violin and a motorcyclist.

What researcher are you referring to? It's entirely possible that only 30 percent of the corpus movement produces audible sound, but the whole corpus surface area doesn't respond equally at all frequencies; if it did, it certainly wouldn't sound like a violin anymore, hence the acoustic filter I mentioned. If this is so, where would you suggest the other 70% of movement is being dissipated? In phase cancellation? Phase cancellation algebraically sums to zero movement at 180 degrees.

I highly doubt Amati knew anything about phase relationships, pressure gradients and frequency spectrum content. He was a woodcarver.

Link to comment
Share on other sites

Mechanical wave coupling to air is certainly not nonsense;

No disrespect intended, Bill. For audio speaker diaphragms with their electromagnets that indeed is not nonsense. We're talking about acoustic violin though - no electromagnet, no RLC oscillation circuitry.

Internal & external pressure gradients around an arched violin shell [from compressed & rarefied air] are all that's required to amplify played string frequencies. Does an airplane wing need mechanical flapping to create "lift" ? [Answer: "No", just a little math; a little geometry :)]

It's entirely possible that only 30 percent of the corpus movement produces audible sound, but the whole corpus surface area doesn't respond equally at all frequencies; if it did, it certainly wouldn't sound like a violin anymore, hence the acoustic filter I mentioned. If this is so, where would you suggest the other 70% of movement is being dissipated?

I somewhat like your "acoustic filter" analogy [for the arched Belly only though].

I highly doubt Amati knew anything about phase relationships, pressure gradients and frequency spectrum content. He was a woodcarver.

That's why I said, "Amati's genius" [someone else].

Jim

Link to comment
Share on other sites

Bill and probably thousands of others basically look at the strings as thee oscillating device with corpus creating mechanical waves coupled to air. Kinda like a speaker diaphragm, huh? That's one way of looking at it. The researchers also suggest only ~30% of corpus vibrations as seen in 3D animations contribute to the sound radiated and heard. REALLY!?!

The other way of looking at it is to drop all this mechanical wave coupled-to-air nonsense which may simply be analogous to distorted lip-flapping by an unhelmeted motorcyclist traveling at high speed. It ain't the lip-flapping that gets you from Point A to Point B [it's those spinning wheels!}.:lol: Let strings rightly be an oscillating device, oscillating 'air' which is compressed & rarefied, then projected to ambient air by pressure gradients while maintaining varying velocity and phase relationships. This is what Amati's genius was trying to accomplish.

Jim

"Let strings rightly be an oscillating device, oscillating 'air' which is compressed & rarefied, then projected to ambient air by pressure gradients while maintaining varying velocity and phase relationships."

I would suggest submitting this to the political party or lobby that you belong to. I think you have a real talent for stringing complex ideologies in a cohesive fashion all the while maintaining the ability to have it not really mean anything. Its really a very rare talent, you have a way of quantitatively easing irrational exuberance with clear and present obfuscation. :D

Link to comment
Share on other sites

"Let strings rightly be an oscillating device, oscillating 'air' which is compressed & rarefied, then projected to ambient air by pressure gradients while maintaining varying velocity and phase relationships."

I would suggest submitting this to the political party or lobby that you belong to. I think you have a real talent for stringing complex ideologies in a cohesive fashion all the while maintaining the ability to have it not really mean anything. Its really a very rare talent, you have a way of quantitatively easing irrational exuberance with clear and present obfuscation. :D

Jezzupe,

Did you at least understand that my second "oscillating" is a verb? :D

Jim

Link to comment
Share on other sites

However, for stringed acoustic instrument math, there are NO imaginary parts. Normal FFT algorithms that handle regular old number theory arithmetic is all that's required.

I think Gabi Weinreich, Joe Curtin's acoustics theory and experimentation mentor, is advocating to use the phase information also from such FFT transfer functions. Using imaginary numbers is a convenient way (for a pro in math) to deal with amplitudes and phase in the same expression.

I think at least data from two transducers is necessary to get meaningful information about the phase. Most users of FFT here, I think, only uses the mic signal or recordings. I think Wm Johnston with his two transducer system could in principle get phase data.

In George Stoppani's software there definitively are used complex numbers (numbers with a real and an imaginary part). A wave (e.g. vibrations in a structure or a sound wave) can be described by a vector rotating in the complex plane with the angular velocity 2*pi*frequency and a fixed phase angle shift in relation to 'something', like another wave, the driving force, etc.

[Edit] I just saw your link to a wikipedia article in the original post of yours, which at a first glimpse seems fine.

Edited by Anders Buen
Link to comment
Share on other sites

The researchers also suggest only ~30% of corpus vibrations as seen in 3D animations contribute to the sound radiated and heard. REALLY!?!

I think it might be even less than that. I think 4%-ish is more like it, depending on where on the frequency scale we are. A friend of mine measured the amount of energy put into the violin that was converted to sound.

I quote the conclucion from one of his conference papers:

A method for the accurate measurement of the radiation and the losses in a violin has been presented. The method can be fast (~15 minutes per violin if a 8-channel system is used). The results will tell which modes are efficient radiators. Further work will include a study of the relationship with the modal shape and the radiation efficiency. For the two violins measured the radiation efficiency in the frequency range 200-3000 Hz was measured to be 1.54 and 1.24 %, having peak values at ~5 %. Since the radiation efficiency is small and depends largely on frequency, measurement of the mobility at the bridge will not indicate all factors affecting the quality of a violin.

Link to comment
Share on other sites

The 'math' behind harmonics is well understood, but not violin geometry. In this case, acoustic violin design IS geometry which effectively IS the "oscillation system". The answer is math/geometry.

It would have been easier to model if the violins were made of the same isotropic material with exactly the same graduations, archings, and geometry. If you by geometry mean just the outer dimensions of a violin, these are of course a part of the story, but not the whole story. Exactly identical violins does generally not sound the same.

A violin is an immensely complex vibrating structure while being played. Not to mention the performer and musical part. The complexity makes it difficult to draw clear conclucions from testing out theories. Building a violin according to a 'theory' and succeed will not prove anything. Buidling one according to no theory and succeed will not prove anything either.

Link to comment
Share on other sites

I"m impressed by the use violin makers make out of analyzers, honestly, Anders and BIll mentioned some very cool stuff, very interesting read, thanks.

Not being a violin maker, but having worked with audio engineering most of my life this all creates a question mark in my mind.

Usually the really musical instruments (all types) and microphones (for me another instrument) have some pretty strange frequency bumps, strange phase cancelations etc, that is what makes them unique and interesting - the flat stuff is dull-boring-unmusical.

When EQ'ing tracks I don't look at what I'm doing, go by feel, when I look afterwards the EQ settings can be quite strange, if I had thought about it I would have never done it.

So for me, the strange and complex curves are the cool ones, the ones that touch my heart, attract my attention, what makes an specific instrument sound fat usually is a strange complex bump in the low end of that instrument spectrum.

The question mark that remains is that - since good and musical is the unpredictable - complex - strange, how can one by looking at analysis know where to go? How to know where and how should all the complex anomalities be?

Plus audio is the same as Joe Robson's logo, adding a little of this changes that and the other.

Analysis are getting advanced but are still blind analysis, every link in the chain is a weak one, for example the converter

192 k = max sample frequency = 96 k = 2 samples of that a second = 8 samples of 24k / second = 16 samples of 12 k per second

All the complex harmonics have gone down the drain - and we are not talking about a timpani but a violin - extremely rich in ultra high harmonics - even those very important ones that we don''t hear - we perceive them. I could go on and on about every link.

"Edit on sept. 5-2011"

"After reading various posts by Anders Bues,David Burges, BIll Yacey and others I have changed my mind - - - my opinion on the subject had been based in the uses of analysis I had seen in my 26 years as an audio engineer-record producer - the analysis here is of a dfferent kind."smile.gif

My link

Please do not misunderstand me I am in no way questioning the value of analysis, it's a very noble pursuit.

Very interesting stuff you guys do, very cool community, I don't now why but I enjoy reading this forum very much, thanks again.

Je ne fait rien sans gayeté

(Montaigne)

Link to comment
Share on other sites

I think Gabi Weinreich, Joe Curtin's acoustics theory and experimentation mentor, is advocating to use the phase information also from such FFT transfer functions. Using imaginary numbers is a convenient way (for a pro in math) to deal with amplitudes and phase in the same expression.

I think at least data from two transducers is necessary to get meaningful information about the phase. Most users of FFT here, I think, only uses the mic signal or recordings. I think Wm Johnston with his two transducer system could in principle get phase data.

In George Stoppani's software there definitively are used complex numbers (numbers with a real and an imaginary part). A wave (e.g. vibrations in a structure or a sound wave) can be described by a vector rotating in the complex plane with the angular velocity 2*pi*frequency and a fixed phase angle shift in relation to 'something', like another wave, the driving force, etc.

[Edit] I just saw your link to a wikipedia article in the original post of yours, which at a first glimpse seems fine.

I understand complex numbers. I get the impression some are thinking of measuring "phase" between what effectively is ALL those herky jerky corpus vibrations seen in 3D animations, or even between ALL frequencies in the spectrum. That may be meaningless. All one needs to do is identify 'where' violin geometry [curves] creates certain phase-shifts [90º for example], then key in on certain frequency ratios with a couple of microphones and maybe an oscilloscope [iF one needs a visual].

Jim

Link to comment
Share on other sites

I think it might be even less than that. I think 4%-ish is more like it, depending on where on the frequency scale we are. A friend of mine measured the amount of energy put into the violin that was converted to sound.

And isn't that enough to discount the mechanical waves coupling-to-air acoustic model? For arched shell geometry, it really should be [enough].

That can't be what the first Cremonese genius had in mind, i.e., he wasn't thinking arched shell geometry would vibrate with herky jerky motion

'catapulting' highly-organized compressed & rarefied air to an audience.

Jim

Link to comment
Share on other sites

It would have been easier to model if the violins were made of the same isotropic material with exactly the same graduations, archings, and geometry. If you by geometry mean just the outer dimensions of a violin, these are of course a part of the story, but not the whole story. Exactly identical violins does generally not sound the same.

By geometry, I mean a very precise 'set of curves' designed to shape & control air vibrations by each bow movement of the Player. Ideally, you want to only model a "perfect geometry" violin, then you understand how modifications to that geometry will affect performance.

With wood being nonisotropic, it's understood even identically-shaped violins will have some differences in resonances. But the resonances important to

controlling compression & rarefied air will be the same.

Jim

Link to comment
Share on other sites

Analysis are getting advanced but are still blind analysis, every link in the chain is a weak one, for example the converter

192 k = max sample frequency = 96 k = 2 samples of that a second = 8 samples of 24k / second = 16 samples of 12 k per second

All the complex harmonics have gone down the drain - and we are not talking about a timpani but a violin - extremely rich in ultra high harmonics - even those very important ones that we don''t hear - we perceive them. I could go on and on about every link.

One thing that may help in this regard is that sound output from a violin drops off severely above about 8000 hz. That's not to say that this region doesn't matter, but that in terms of listener perception, its the overall sound level in this region which seems to matter more than the detailed frequency recipe (for violins). Exceptions might come up when an extremely high note is played, furnishing the lower and more musically familiar end of the harmonic series, but playing super high notes is usually a very minor part of listener evaluation of overall violin sound quality.

So what I'm trying to suggest is that with all the inaccuracies of acquisition and processing, there still seems to be enough information to reproduce violin sound in a way that it tracks pretty well with our perceptions of live violins. That suggests that a lot of meaningful information is there to be extracted.

Jim, I think your ideas might work OK if violins didn't change with time and conditions, and if violinists only needed to play a few select notes.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.



×
×
  • Create New...