Bill Sampson Explains Violin Building

[@kenstevens4175]

Fantastic explanation. 👍

[@beverleysampson6354]

Thanks for the kind comment Ken. This video was done about 13 years ago and seems to be still circulating on the Internet. Glad you found it. Billy

[@trailkeeper]

@WGSViolins Some computer software you can download for free can generate (audio) signals at frequencies the user can set. Probably have to use some speaker connected to the audio output of a computer to hear it and for the wood to vibrate to that.

[@Satchmoeddie]

Respond to this video... I have to say thanks to the maker of this video, as I know how he likes to see the back and top with a powder media vibrating on them. Everyone is different. I did a DIY thing from an old AE magazine that allows me to find a top's resonant frequency. You need a speaker and a mic, or piezo pick up. Either will work. When the signal get larger and you can stop it with hand pressure you have some resonant freq, (largest is primary) I'll try to find which mag #.

[@Satchmoeddie]

@WGSViolins Low freq ones are on eBay for decent $ but you need the counter too, so I suggest one made by B&K that had a freq counter built in. $300-$400 new at Fry's Electronics and it has a good output. It is great for amps, speaker testing and resonance and this trick as well. For another $50-$100 Jameco has a neat one with a fine tuning knob that I wish I had bought. It is also a B&K (not to be confused with Danish made Bruel and Kjaer the Stadivari of audio test gear since the 30s)

[@Satchmoeddie]

@Satchmoeddie In the DIY resonant freqency finder fo speaker cabs, or other items an O scope, freq gen, amp, speaker, mic and Z input are used, so hopefully you can do the lisajou display. It is not crucial, but it helps. I will try to find that article, and and post it on my YT or other place with a demonstration. It is like fancy tap tuning or an addition to tap tuning and this method. An octopus box may also work for a display of the resonance into the mic and back to the scope. Nice Traynor

[@Satchmoeddie]

@mechmusicman You hope like hell your glue job and the kerfing or inside gluing strips are also symetrical with each other, or it will be off. This why he said it should be not that it would be. Wood being a natural material can have some flaws in it, and sometimes glue wont take as well in one spot. A sound post is also set between the two plates, and it's positioning is crucial too. It transfers string energy to the back plate. . Pardon my spelling. I put in a new graphics pgm. Print shrunk!

[@bubayou]

It is called the Genius Ear. If you have perfect pitch all you need to do is to tap the plate and you can tell the resonant frequency.

[@Satchmoeddie]

@Satchmoeddie I use an old one for speaker voice coil rub and audio made by Beckman I got for maybe $50, but its output is low, at < 2 Hz the sine wave is sloppy, but with an amp it does this with sine, square and triangle wave forms. The .1 Hz is perfect for detecting speaker voice coil rub, and with the B&K's counter I can dial in audible stuff very well as it has a 5:1 knob that adjusts very slow and fine to the frequency reqd. I started Luthrie in 1983, but I'm in Arizona, so I do amps too.

[@brangreen85]

How did they do this in the days when this technology was not available?

[@chaddonschaddons7084]

Important questions have yet to be answered. Why does the shape of the violin have to necessarily be that shape? It is a design from a few hundred years ago and has been slightly tweaked through the years but certainly a different design could produce a similar sound yet slightly better whether audibly and/or aesthetically. The complexity of the learning curve and playability is a factor called into question for the need of redesigning it. I have many questions at an engineering level that I doubt in my frustration that anyone can answer without being dismissive as to say, "Well, that's the way it's been done therefore although your design might sound and play better regardless, we have to stick to the old design because yeah". The backbone for my concern is the resonance and radiation of the sound through the front plate and f-holes. The shape of the patterns for their respective frequency responses don't radiate through the instrument optimally in its current form and there is so much loss that to me clearly sounds like a cat muffled inside a 1/16th veneer laminate board box.

[@beverleysampson6354]

The current shape of the violin resulted from an evolutionary process rather than an architectural or engineering one. It began with a rectangular box whose corners were rounded for comfort and whose center bout, or waist, evolved as a result of its bow striking the body when reaching for the outer strings without striking the inner ones. Further evolved features were mostly aesthetic, such as the shape of 'f' holes and scroll as well as the smoothness of finish and varnish color. Larger concert halls of the 18th and 19th centuries required more output from the instrument and this marks the first time that acoustic investigation and engineering demands prompted revisions to its structure -- thus, the neck was tilted back. The tone was, more or less, accepted by players and audiences. It wasn't until much later that its tone was being investigated from the viewpoint of improvement. So, centuries of familiarity, by players and listeners have applied so much inertia to the notion of changing anything about the instrument that it was not until the past hundred years or so that changes were even considered. My belief is that if an architect, acoustician or engineer did produce a striking improvement in any quality of the instrument, whether sound or playability, it would be accepted -- perhaps slowly -- but eventually.

[@potaylo]

would the bass bar change the symmetry and resonant frequencys of the top palate?

[@WGSViolins]

Hi Pedicravo, I'm not sure of an absolute answer to this question but do note that the plates are sectioned & each section is a different size. It would seem that each would be responsible for amplification in a different frequency range. When you include the bass bar distributing lower frequnencies over the length of the plate the whole then acts as a giant frequency filter/amplifier. There may be a better asymmetrical Chladni patterns for tuning but it seems research has not yet found them.

[@bubayou]

Have you experimented with trying to control the Q AKA Bandwidth of the plates If so do you have an opinion on what would be a good target. Or is it function of the quality of the wood

[@dadnapt24]

how did they do this in the 18th centry?

[@WGSViolins]

I think the best that I can do at this point is refer you to an article that appeared in Scientific American, October, 1981 entitled The Physics of Violin Plates by Carleen Maley Hutchins. She is the co-developer of the electronic plate tuning technique and she explains in great detail how the method came about and how, over many years of trial and error testing, by taking apart good sounding violins as well as bad, they arrived at the procedure now in use by vioin makers world wide.

[@jmarett22]

Bill, are there published frequencies for the cello?

[@KellyMercerhfx]

Nice work!

[@jmarett22]

Thank you, I have purchased all of the equipment to simulate what you have done. I'm working on an old cello I acquired and hope to help it out.

[@pipesmokingbearguzzlingbas960]

Alexandre Marques Bemquerer If you look at old cremonese violins most of them were indeed assymmetric, both in plate thicknessing and overall shape (obviously not by much, but by some margin), and some people speculate that that might indeed have been completely purposeful.

[@WGSViolins]

@Lafiddler1 I'm not sure I can answer that directly. Mine is an old Hewlett-Packard analogue model 200CD from the 1960s which was given to me by a former student of mine who was replacing laboratory inventory in the 1980s with newer digital models. Some of the electronics supply stores such as The Source may carry them. Try "Fluke" on the Internet. Good luck. Bill Sampson

[@WGSViolins]

I use only the 3 octave frequencies recommended in the Scientific American article by Carleen Hutchins.

[@pedicravo]

its seems like the plate is up to display a symmetrical pattern if its been symmetrically thickened out, whether be it a good or bad thickening way, i don't see how this method could really tell a good from a bad violin. and, considering the wolf tone, that is caused by resonance issues, maxing out a frequency response is more likely to lead to wolf notes isn't it? wish i could see a tuned violin by this method, to be able to evaluate it

[@WGSViolins]

jqsmith2011, Sorry for the delay in responding. I don't check these posts often any more. The researchers (Carleen Hutchins, John Schelling & Fred Saunders, ca. 1950s 60s) worked backwards from high quality violins to poor quality violins -- more than 500 -- in attempting to establish frequncy values for good quality instruments. They arrived at the current optimum values by testing the free plates of those instruments, tabling, averaging & using them on violins under construction. Bill

[@WGSViolins]

Thanks Kelly. Bill

[@airplanegeorge]

cool stuff but how the hell do you play it?

[@WGSViolins]

Steve, It will be playable. Some better players may not think it tonally balanced or not responsive enough but most (even good) violins will elicit some criticisms. As long as it meets the specifications set out in the instruction book all you need is to ensure that it is set up properly -- that is the bridge is properly cut, fitted and sited, the strings are properly spaced and their clearance over the nut is proper: the same for the soundpost. Exercise patience in all of this. Bill

[@pedicravo]

hi bill, considering that the inside of the violin is setup asymmetrical, has asymmetrical force load, and that the strings vibration amplitude and placement is also asymmetrical, i don't properly understand why we should go for a symmetrical vibration pattern, specially when we consider that lower frequency play better on a bigger speaker and a high on a a smaller one, can you give me a light on that?

[@nckeller]

Physics is a wonderful thing. I imagine you could learn the sounds and accomplish a similar result by tap tuning the plates. Make a couple hundred of anything by hand and you probably would get the hang of it. Even some of the early masters works came out much less than stellar.

[@WGSViolins]

The Catgut Acoustical Society. Check "Scientific American, October, 1981, article by Carleen Maley Hutchins, The Acoustics of Violin Plates. Empirical research on some 200 violin plates showed that (paraphrasing) the best quality violins had free plate vibrational modes or tap tones around 360 Hz.

[@pedicravo]

just read it, good article, thanks for sharing, helped out understanding the method and how it works. but even in the article, if u check the response graph of the violin they made ( when compared to a guarnerius) you'll see that the instrument is no responsive in some frequencies, which probably is for wolf notes. and although they claim the instrument to be good, a instrument that has many flaws in the frequency range is surely no perfect thing. thanks for the chat bill, was instructive.

[@scody311]

For my first violin, if I don't have a mentor or a teacher, and am doing this all from a book... and don't tune my violin plates per your demo here... will the violin be HORRIBLE? or... will she at least be playable?

[@potaylo]

What are the 3 frequency’s or are they different for each violin?

[@briteenough2burn]

The goal is not to measure frequency response, but rather to USE frequency response to carve the top and back plates evenly. Assembly may change the frequency response but it certainly doesn't change the physical characteristics of the wood itself. Nice try, though.

[@SerpaKerphia]

Stradivarius would have used a set of tuning forks, slower, but the same principle applies.

[@tonytrilex]

cant you just use a chromatic tuner instead of all that

[@rhykko77]

lol good one !