Anders Buen

I have a Hypothesis that lower arch may lead to less movement from humidity variations. A flatter arch makes the plates less stiff, and they may need a thicker construction then (like for many del Gesus), to act stable, both to climate change and to string pressure. Some fine instruments can have a negative arch in the middle, like a "shallow saddle". It is all about balance, I guess. A safe way to deal with wooden musical instruments is to keep them in a controlled climate, as they do in some museums. E.g. 40-60% or even narrower ranges if possible. They do for the Viking ships, sledges and objects at the Vikingship Museum here at Bygdøy now in the new building opening in 2028. And while storing the objects during the construction work (they are stored in the old museum building next to the construction site for the new building) I have also recently learnt that in the Munch paintings museum, they control the climate to reduce the stress on the partly brittle paintings. Edward Munch, the painter, stored some of his works outdoors, in rain and all kinds of weather for a while. So these need less "aggressive weathering".

Deformation will also appear from humidity variations from the climate, as the dimension change and changes of elastic parameters and damping is different in the principal directions along (along green tree direction) and across the grainlines (radial direction) in the wood. It is largest in the tangential direction, in the «thickness direction». The dimensions of the directions, along with the %swelling or %shrinking from the MC content movement and elastic change, will be the driving force, but some restriction of movement also probably happens, which may result in internal stresses, and possibly cracks. At higher humidity levels and resulting moisture content in the wood, the wood will be more flexible and also more likely to creep under the load. Creep is deformation that still will be present after the construction is no longer under load. Part of the «Cannon study» is using a «mild» RH change and monitoring the resulting deformations. The natural variation in the local climates may be much larger, and at RH above 60% the creep risk increases faster than below that «limit».

I have colleced and averaged many thickness datasets into one with spread around the averages. However I am not so sure that the graudation pattern matter so much, more the average thickness. I have shown that using Hutchins tap tone advice from the 80ties most likely leads to too stiff and heavy tops. Most likely they were "inverse- inverse graduated". That is thicker in the centre than in the upper and lower bouts a classic violin making trait, until more became known about the Strads and del Gesu violins. Many violin do get patches. I have had to patch up two of my instruments and they did not sound much different afterwards. The extra stiffnesss stbilised the plates anough fot them to work through the seasons. Many fine violins do get patches, strips and edges doubled or new bass bars. It may or may not influence the sound. I have no experience with these inetruments pre- or post refurbishment. Have seen that the sprectrum of the Titian Strad taken by Saunders with his bronze wire in magnetic field method from the 1940ties does not look similar to todays more recent tests. Could be the metod differences, but may also be related to acoustical changes. Setup is also in that category. If the graduation patterns were so important, why would adding patches be used as a repair? I do not know how common it is. I have several proofs to back my intuition which works best for plates you do follow from start to the end, that is with the same wood. So I found clear correlation between the plate tones, weights and graduations of Martin Schleskes Master theses work from Müller BBM, even if he claimed there were none. I probaly have better skillls than him, at that time, at finding any correlations. And he focussed much on mode 5 which is one of the weaker ones in that set. Don states that he see no correation in his dataset, as I claim. I have part of that dataset and have put it into a larger collection showing very clear (significant) correlations between free plate mode frequencies and the signature mode B1- and B1+ mode frequencies. There are variation, we may easily connect that to wood properties. But as Sam Z and co workers point out now, models may play a significant role. That may also include choice of wood for the models. (Del Gesus lighter wood, Strads vice versa) The is no doubt that the correlations are as I say. If they do not show up, then there must be input parameters that are not accounted for, like wood properties, a usefaul parametrisation of arches, sizes, and so on, factors that may carry more information. I was one of the first, if not the first, to investigate how graduation patterns may influence sound. It stated with the Miracle makers later on my own worshop set. The workshop set correlation matrix is shared on Researchgate. It was a 10-12 year work. However the soundcard for the impact hammer was not stable enough so I could get different readings on the same instrument on different days on the v/F level [dB]. But there are data from playing scales as well in it that may be more stable. I also have a larger dataset of plate data from the literature, e.g. the dataset from the Bibao experiment, where everything lines up, because the wood is chosen to be similar. I think I have shared that too here. In this bigger set I try to add all relevant material parameters, arch height thicnkesses, wood density, E-moduels and G modules. Tha aim is to use it for an improved model based on Morals work I published a draft for in 2002. If I cannot get a working statistical se out of it, I plan to fill in with FEA calcualtions. When you thin a plate, it becomes lighter, less stiff, more flexible, and in general the trend for the mode frequencies are that they drop. I think the graduation pattern is not all that important. Certain areas need to be thicker, like the block regions, in front of the f-holes and to the sides there, partly outside the f-holes. F-hole wings should have a decent mass and strength. The corners in the upper and lower bouts can be thin. The bridge region should not be too thick.

In Martys given plate working strategies, they are all related. The thickness, the plate weight, the impedance, the tap tones and the stiffness are all related fairly closely. Does not address the distribution of thicknesses, but still. I guess the stop criterion may differ for the different strategies, and may turn out to be the main influencing factors. If you work with larger datasets of these input parameters for violin plates, it is quite easy to see these more or less close correlations.

Much looks “cheap” on this violin. The top wood plate wood, the plain back plate wood, the plain and, likely spirit based varnish, the color, the finishing of the scroll etc. Usually fast made mass produced instruments are thick, as that is the safer choice against cracks or creep. Average thickness around Strads near the f-hole is about 2,6mm, and up to 2,7-2,9 mm outside a bit south (avg of 70) Del Gesus: A little thicker, the wood are also in average lighter and probably somewhat less strong, 2,5-2,8mm in average inside and 2,9-3mm outside. Average of 50 instruments. Fine violins get their edges doubled when the tops have been off and on too many times. This may influence the border thickness to some extent, I guess. The wood may be heavier than ideal in this violin, then it may be stronger, but not if the grainlines lie at a different angle than 90 degrees to the rib plane. Stiffer wood needs to be thinner to get the best out of it. However, there may be many possible usable solutions.

Could be that the A0 (main air mode, f-hole mode) is high and couples with the string. Del Gesus have a higher A0 frequency or tone than Strads, and in analyses of strong G- playing (start of the Sibelius concerto solo part) the D starts to ring. Ringing strings is a sympathetic effect. If you close one f-hole partly the A0 will move down in frequency. Does the rigning disappear then?

Still thick, but know it can happen.

4 mm in the central region of a violin top plate is thick.

There is a profile on that name on Facebook. Any decent name is accepted there. On education it reads: Studied Violin Making and Restoration at Newark School of Violin Making I know there are many MNers who did study there. He looks a bit older than those I know were at Newark. My grandmother became a school teacher at 50. An exotic name, for sure!

My small experience with slab cut backs are that they appear to give violins with a low frequency B1+ resonance mode. As the hygroscopic shrink and swell directions of maple is larger in the sideways direction of the plane of a slab cut plate (the tangential direction) than the standard, radial direction. Possibly a larger risk for cracks, and a somewhat "softer sound".

I find this tread intriguing. I'll just share documentation of a test I did many years ago (2008) using putty on the f-hole wings. I see Don has made much more into the analysis and I guess we see similar results. 120426 Mass on inner f-hole wings shortened.pdf

Erik V Jansson did experiments with stiffening and cuts near the f-holes with the Polish violin maker Niewzyk many years ago. Thy stiffned the region in front of and behind the f-holes in an attempt at manipulting the bridge/body hill of the instrument. It is probaly best not to use any stiffener. However, I have some violins made of stiff top plate wood, also stiff in the crossgrain direction. They are loud, but maybe not with the sweetness one could prefer in a fine instrument. Several makers point toward a low crossgrain sitffness as being beneficial for tops. Wether it is true and working or not is hard to say. But I believe many fine violins do have off 90 degrees grain angle to the rib plane, possibly at the cost of a little less power and punch.

I do not know. I have heard about small braces used to stiffen up a too thin border region in violins, from a good violin maker with restoration practice. Possibly an idea by one of the "main restorers" that also attended the Violin making workshop at Oberlin one year. If the radiation efficiency thing is correct, a plate with stiffeners will radiate more sound than a plain one. However, the braces add weight and stiffness, and reduces the mobiliy of the plates making them vibrate less for the same energy input. The balsa violins by one of the Oberlin attendees (I do not remember his name right now, but know how he looks, - sorry!). He used a brace pattern on the outside of the bodies. Even if they did not look like a violin, they did sound like one. Much is of course in the bowed string part, but still there is a "filter" in the body radiating a given "voice".

There are a mix of mode 2 and 5 in the B1- and B1+, just opposite in the two cases. Not the same, but some resemblence.

Open seams will probably affect many resonances. However I think the effect may be smaller on a really stiff instrument than on a more flexible one. Sometimes a good glue seam seem to improve instruments from its former less well glued state.