Sign in to follow this  
Oded Kishony

What's wrong with this picture?

Recommended Posts

I watched the short interview with Sam Zygmuntowich where he talks about the 'invisible violin' What I think he means is that various analytic methods have revealed aspects of the instrument and how it works that are not apparent by ordinary means. Sam says in the video that having a realistic mental picture has helped him make better violins.

 

I don't doubt that. However is the picture presented by modal analysis the real picture?

 

I don't think so. At the very least it is only a part of the picture. 

 

Some observations:

 

>>M.A. does not make any distinctions between those resonances that radiate to the far field and those that don't. According to Oliver Rogers only about 25-30% of vibrations result in far field radiated sound.

So at the very least the picture you're looking at has a lot of 'static'.

 

>>Modal analysis shows the movement of the surface at distinct resonant frequencies. Some of the intervals between modes is quite wide. There is no information about what's going on between frequencies

 

>>M.A. relies on a computer generated sound to produce clear images. It is quite a different story from what happens when a bow scrapes a string.

 

>>M.A. implies a static situation, but modes are NOT stable, they shift in frequency and amplitude depending on bow pressure, hand pressure on the neck, shoulder & chin pressure and type of shoulder rest ( or not ) used.

 

>>Other 'scientific' investigations are also oversold, in my opinion. For instance UV photos of varnish yield almost no useful additional information since a combination of ingredients can be made to fluoresce in a practically any color. At best perhaps they eliminate an ingredient or two.

 

>>While I concur that having a more correct mental picture is better I question whether the effort, and steep learning curve required will ever pay off. At the end of the day Modal analysis has almost no predictive value. No more than a well qualified experienced luthier.

 

I'm all for robust scientific exploration of the violin, I'm dubious about the current approach and tools being employed.

And I'm waving a flag of caution, what's wrong with this picture?

 

Oded Kishony

 

 

 

 

Share this post


Link to post
Share on other sites

Oded,

 

I do not think Sam Z. was referring to modal analysis. In any case I do agree with your concerns; however, I think you are obfuscating the concept of controlled scientific experiments. This is what Sam Z. was discussing. They don't always simulate ideal realistic conditions, but they are repetitive enabling examination of different conditions and their results while constraining other effects.

 

Mike

Share this post


Link to post
Share on other sites

Just some corrections to start:

 

Computer generated modal analysis does not use computer generated sound. It uses an elastic model of the plates to compute "resonant" deformation patterns through a mathematical analysis that searches for "eigenvalues" and "eigenvectors". More on this is beyond the scope of this forum.

 

It is a firmly established engineering principle that ANY "linear" deformation of an elastic solid or plate, static or dynamic, is some sum of these modes. It is even possible to account for important effects like vibration dampening.

 

By looking at the deformation pattern of a mode, you can usually predict if it is a sound radiating pattern or a non-radiating pattern. Patterns that are symmetric about stationary modes or have an odd number of peaks tend to move air that can be heard. Asymmetric /even number of peaks will generate sound waves that interfere/cancel, i.e., non-radiating. It is important not to confuse this with frequency based radiation of sound through air. Once a radiating sound is emitted by the violin, other physics comes into play as to whether or not it will reach the back of a concert hall.

 

That said, I will admit that just looking at the first several modes as static deformations will give limited information about transient behavior. But they can be excellent predictors about sustained behavior, like predicting potential wolf-tones, the influence of bassbar designs and soundpost locations, or maybe the effect of various arching or plate thickness patterns. The frequencies of the modes and their relative strengths, typically called the sound spectrum of the violin, can be a strong predictor of tone potential.

 

I have used transient modal analysis extensively for non-violin applications. Parametric studies are very helpful in both informing you about what things you can change to actually do something useful, and for dispelling common myths about how things work. But I have never seen anything extensive done for violins. Poor Stradivarius had to hand build a full violin over a month using valuable wood to test ideas. If we were motivated enough, we could try out changes in mere hours using modal analysis.

 

Of course, you would still have to build something to hear how it sounds. But as my father once told me, "A journey of a thousands miles begins with a good map!"

Share this post


Link to post
Share on other sites

Mike's dead on.  What you record in the laboratory is what you can do and are able to observe in the laboratory.  I'll note that papers where computer simulations are being treated as data always make me a little nervous, as well as that conditions at discrete modal frequencies are going to be dynamically added by something called superposition.  Still, any insight is better than none  :)

 

Anyone desiring a better insight into discussions like this, which frequently happen here, is encouraged to buy or borrow this book which will be a fun read, explain a lot about wave functions and painlessly teach you basic calculus along the way.

Share this post


Link to post
Share on other sites

What's wrong with this picture?

 

Nothing is wrong with it.

But it's pretty much an analytical point of view, and of analysis. Which is how the population today was taught to think and to *analyze* any and everything (or every problem) in life.

So, from that aspect, or from that point of view, what he's doing is what is expected. 

He's absolutely correct that a good sounding violin is possible (perhaps, as he says, is even relatively easy) to make, but that reaching to the last few percentage points of "sound" becomes more and more difficult.

 

He's also correct that the real contest is to arrive at that result. No matter what.

That neither he, nor anyone else I know, has REALLY arrived there yet is almost incidental - because the mechanical, or perhaps it is more correct to say that the "correct mechanical" requirements have not (in my opinion) been re-discovered yet. Nor have they been re created. I haven't listened to every makers new violins, so, anyone can counter what I'm saying and most likely will.

Oh well, so it always goes.

That there exists a system for making violins that takes into account all of their different variables, and combines them in some correct way, should be almost a given. I believe that such a "formula" exists and that it is beyond us today in general to embrace, or to practice - for whatever reason.

 

But...that this, or that such a methodology is still hidden today, for some (whatever) reason, is also a reality that exists currently. In relative degrees - I agree. But that the exact formula (if it can be called that) is not readily apparent. Nor is it a 'secret' known to only a few luthiers.

 

I enjoy the fact that Sam is not mysterious about the fact that we are still looking into various activities that may shed some light on exactly what it was that those few luthiers of old were doing, so some of the mechanical aspects of it, may eventually be re-discovered and then applied generally.

IF - age itself, combined with adequate making skill, isn't the 'missing ingredient' for the top tier in violin making today.

I do often wonder about that one.

Because nothing but time will shed any light on that.

Share this post


Link to post
Share on other sites

I believe I've gotten a lot of value from the various violin body excursion simulations and measurements, dating back at least 20 years. It can help link sound to the realm of the visual, and I think we're better off if we can bring both into play.

 

Still,  as Oded said or implied, jumping to any conclusions can be treacherous. There are some motions on violins which don't appear to do much towards emitting sound to an audience. Still, that won't mean that they don't matter in terms of feedback and playability to the player.

 

Sam is a fairly analytical guy, and also a pretty successful maker, so maybe there's something to be learned from that.

Share this post


Link to post
Share on other sites

I believe I've gotten a lot of value from the various violin body excursion simulations and measurements, dating back at least 20 years. It can help link sound to the realm of the visual, and I think we're better off if we can bring both into play. But as Oded implied, jumping to simple conclusions can be treacherous.

When is the book coming out, DB?

Share this post


Link to post
Share on other sites

When is the book coming out, DB?

LOL.

Julian, we've met and talked, and I happen to think that you're a really bright and perceptive guy.

 

I struggle a little bit with the concept of "raising the bar" in violin making. Is it needed? Probably not. We're basically working with ideals which were loosely established 300 years ago. Really good musicians can do a pretty decent job with whatever they have to work with. But they are also highly perceptive and appreciative of small improvements.

 

Do I have some tweaks which I think will help, in a highly competitive environment? Sure. The question becomes what to do with them. Do I post them here, thereby giving them away to Chinese manufacturers? I struggle with that. Loved the Chinese people when I was there.

 

I also think that something around 5% of contemporary handmade violins are pretty darned good already. Maybe that's all we really need?

I don't know. What do others think?

Share this post


Link to post
Share on other sites

 

I'm all for robust scientific exploration of the violin, I'm dubious about the current approach and tools being employed.

And I'm waving a flag of caution, what's wrong with this picture?

 

 

It's mostly a marketing tool.  Useless for violins though as all one gets is pictures after the fact. :)

Share this post


Link to post
Share on other sites

First, let me just say I like Sam, and he's not only a maker with fantastic tradtional skills, but also a very articulate guy.

 

But, as far as I know, he does not actually use any of this newfangled technology in actually making a violin.  Last time I quizzed him, he didn't even measure wood properties; he just judged wood by how it responded to handling.  That's not necessarily a criticizm.  Although I do measure wood properties, modal analysis does not come into play when I actuallly make an instrument. 

 

In that regard, I'm very much with Oded when it comes to questioning what modal analysis can do to improve things.  Mostly I have proven to myself with modal analysis that many modes seem to be quite resisitant to change, thus modal anlaysis doesn't help much.

 

 

 But that the exact formula (if it can be called that) is not readily apparent.

 

It is readily apparent to me that there are many different "formulas" that work, in that there are many different-sounding violins that are quite good, each with its own, different arching and graduations.  Unfortunately, the visible, measurable parts aren't the whole formula... there are dozens of wood properties variables (or infinite, if you count variations with frequency or location) in the formula too, and nobody knows what they are. 

 

The key concept I think is that the modes and response are totally determined by the shape and wood properties... so if you can duplicate the shape and ALL the wood (oh, yeah... varnish too) properties, the acoustic result (and modes) will have to be identical.  The problem is to get the wood properties the same... which is virtually impossible, given the number of variables involved.  It is tantalyzing to think that with the modal analysis of a great instrument, all you have to do is match the mode shapes and frequencies, and you'll have the same thing.  I don't believe it can work that way, as most of the variables are locked in place as soon as you pick out the wood to use... and it won't be the same.  (This is one of the reasons why tuning of the plates of the finished instrument I don't believe can get there). 

 

Most of you know that I've been experimenting with themal processing to modify wood properties, somewhat as a form of accelerated aging, in order to see if that is the key to the sound of great old instruments.  While I do believe that the results have been better than unprocessed wood, there are still some tonal features that more resemble new instruments than old ones.  Up until now, I have primarily focused on longitudinal properties, but I believe curious1's comments about crossgrain properties have a lot of merit, and I'll definitely be looking in that direction. 

 

Getting back to modal analysis... I do like to know mode shapes and all that, even if it has had hardly any benefit to making.  I think it has had a tiny bit of benefit to thinking about soundpost adjustment, and perhaps just a little influence on my arching and graduations.  Nothing big or important... but maybe a little.  It is easy to get caught up in measuring modes and trying to control them... not a good investment in time, IMHO.

Share this post


Link to post
Share on other sites

What do you consider to be evidence?

  

Well, if we are talking about "going way beyond that" and the 'that' we're talking about is making violins, that are consistently better or superior to the readily available violins of today, using such scientific exposition as "modal analysis"... then the evidence would be in the product(s).

The violins(s).

 

I'm suggesting that the evidence (or, perhaps, that such evidence) is simply not there at this point in time. 

 

Of course, then again, this is a present day discussion that really has no successful conclusion, because both sides of the basic issue here will pontificate on the finer points, until the end comes nigh, won't they?

Share this post


Link to post
Share on other sites

What's wrong with this picture?

If you've been using a dull knife to cut an F hole you would not know it until you've used a sharp one.

I think it's an emerging realization that M.A., used solely for analysis, is a very dull knife indeed.

The status of violin acoustical research today is up to it's proverbial axles in mud. IMHO.

Oded

Share this post


Link to post
Share on other sites

  

Of course, then again, this is a present day discussion that really has no successful conclusion, because both sides of the basic issue here will pontificate on the finer points, until the end comes nigh, won't they?

Don't know who you're talking about, but it couldn't possibly involve me. My emphasis is clearly on porntificating. :lol:

Share this post


Link to post
Share on other sites

I think the glitter bouncing around on plates into predictable patterns at different frequencies, is fascinating. I paid attention to the Hutchins stuff. But I don't know how a luthier can use this information to actually make a better instrument. The one who claimed to be able to (physics prof who toured with violinist Rosemary ...whatshername) was not doing anything that seemed, to me, based In either science or reality. The demonstration I saw...um...just bizarre.

Basing everything you do with plates off of some ideal frequencies/patterns of glitter (lol) reminds me of psychics who read tea leaves.

Not sure I have ever read a convincing article on the usefulness of some of the tech stuff. I would like to. Links would be appreciated. I truly am interested.

Share this post


Link to post
Share on other sites

 I truly am interested.

 

And I believe you.

Interested in what, though?

You don't make violins, so you're an interesting "intruder" in this.

But not an intruder in the bad sense, but an intruder with a genuine interest.

Are you interested in the theoretical only? Why don't you make a violin? are you simply not able to take that final plunge into the realities you speak so knowledgeably about? 

Hey, I'm just curious.

Share this post


Link to post
Share on other sites

I want to be really knowledgeable, and I'm not. I'm not sure if you really believe I am, or you are baiting me again. Hmm? Well really im not. I learn more all the time about all of this, and it is all fascinating. I like doing little things in the shop for now.

I'm not good at projects that take forever, or especially that would require me to be super meticulous for such an extended period of time. There, since you asked. Once I know how to do more things, and once i have learned more how to see and discern these things, maybe then.

Still not seeing article links.

And I don't know the difference between plate tuning and using modes to build a better instrument. Isn't there a difference? Is that what is being discussed? Someone tell me. Lol

Share this post


Link to post
Share on other sites

 

And I don't know the difference between plate tuning and using modes to build a better instrument. Isn't there a difference? Is that what is being discussed? Someone tell me. Lol

 

 Well, you seem to be asking some of the correct questions.

 

That's a great start. So, you can join the crowd.

But if you never build an instrument  - you'll never know what to ask, really, or why to ask any valid questions, period.

And the questions that you should ask, may be the result of the ability or inability of the violin that you make, to respond in the manner that we ask violins to respond to today, to satisfy the needs of (rather advanced) musicians.

 

Simple, right?

Share this post


Link to post
Share on other sites

Luckily my husband makes awesome instruments. So I'm asking for me, but also for him, maybe to help him open his mind if he needs to. He thinks the mode stuff is not something to pursue for himself. I saw some video of this certain tuning luminary before a live audience that was so shockingly, hilariously, mind-numbingly idiotic that i was agreeing it is all stupid. Good for my husband, i thought, keeping his hands and ears tuned instead, etc.

But some of you, great minds, seem to be convinced of this tech stuff or you have been. I want to know why. I have much respect for so many of you people. If there is real science behind it, let's use it! I understand the need to quantify what was done, and it sounds great--follow a formula.

Where is this Sam Z interview? He wrote a great paper on arching that focuses on understanding the theory. If he has switched theories, I wanna know. My husband would want to know.

Truth is, I am on MN because my husband works all the time. I get bored after work (during work) and this is fun for me since box wine is too much of a luxury right now. Tmi...I am also here to get information which he otherwise wouldn't. Simple! If I have given the impression that I like playing around or bickering more than I like getting actual information, sorry. But...I also see what David is saying about not screwing the careers of every one of you by publicly posting what is actually known, by some.

Does someone have useful links? Please?

Thank you for the book rec, VdA! That's a great start.

Share this post


Link to post
Share on other sites

I think what he said, which is what I have said before :D , is that "strad 3d" helps him envision/think about the way the violin moves. When I look at the undulations I am not correlating that with frequency or even sound, I'm looking at it for what it is, movement.

 

If we take any moment in time, say 5 seconds of movement where say the undulations are making the plate twist, and the lower right bout is high and the upper left is lower and then they reverse, I am looking at the peak "throw range" and or how far does something twist or move one way or another, keeping in mind that the animations are exaggerated.

 

All I know is that my violin making and all that I have learned from it helps with guitars, even if the string motions are different, there are many things that cross paths, and as always wood selections are paramount imo.

Share this post


Link to post
Share on other sites

  

Well, if we are talking about "going way beyond that" and the 'that' we're talking about is making violins, that are consistently better or superior to the readily available violins of today, using such scientific exposition as "modal analysis"... then the evidence would be in the product(s).

The violins(s).

 

I'm suggesting that the evidence (or, perhaps, that such evidence) is simply not there at this point in time. 

 

Do you think that contemporary violins are all pretty much the same?

Share this post


Link to post
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...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.