7 mins
Set the records straight
Acoustic testing can generate a huge amount of data on a single instrument. Rainer Beilharz shows how to record it in a meaningful way for visually oriented people
Lutherie
A violin held in a simple acoustic testing rig
ALL IMAGES COURTESY RAINER BEILHARZ
It has become commonplace for violin makers to collect at least the basic acoustical data of the instruments they make. In some cases, new data is taken from one step to the next – aprocess that can create a daunting amount of numbers, especially when they are on a computer spreadsheet. These may be wonderful for analytics, but I have found them impractical in a workshop situation. It’s easy to lose your place when reading numbers along a line, especially when you are leaning across a workbench to scroll across a large spreadsheet. Numbers on a line can also seem very abstract and hard to put into context with each other.
My preference is to write down acoustical data and other notes by hand, for a number of reasons. Handwritten notes are more practical on a workbench; they’re easier to be spontaneously added to; and they are more memorable, which is useful if you wish to keep a number in mind for reference as you work.
Some years ago I decided to try to collect data in a more visually oriented way, and a very simple sheet has evolved into a three-sheet system covering the making of the instrument and some sound assessment at the end. In writing this, I am assuming the reader has a basic understanding of violin acoustic terms, as an explanation of modes and reference numbers is beyond the scope of this article.
Sheet 1 is fairly straightforward. I keep minimal thickness data, just enough to give me a picture. There are some simple wood properties recorded here (D = density; S.S. = speed of sound along the grain), and I try to make a judgement on the inherent damping of the piece of wood. As you can see, modes 2 and 5 are placed on simple scales marked in 50Hz intervals. Having the scales alongside each other helps me to see the impact of adding the bass-bar and adding the ribs in context. There is a slot for some basic bass-bar measurements and a note on the basic shape of the bar, and I also record bass-bar stiffness as a percentage of the final Mode 5 frequency. It’s not the whole picture, but as a reference it works for me.
‘Harris’ refers to the Harris number, a calculation combining modes 2 and 5 and the weight of the plate, which creates a reference number describing the resistance of the plate to the input of energy. This calculation, originally proposed by Nigel Harris in his 2005 research paper On graduating the thickness of violin plates to achieve tonal repeatability (bit.ly/3w30WbU), doesn’t take into account the damping of the plate and I simply allow the number to creep a little lower if the wood seems to have strong damping qualities or vice versa. There is a spot to record the Harris number for the back as well, but I personally haven’t found it as useful for violin backs so I leave it out, because I don’t want the distraction (although I do include it for cello backs – there seems to be a stronger correlation between the Harris number and responsiveness in cello backs in my observation). I can still work it out from that data later if I want to.
SHEET 1 Instrument data including modes, thicknesses, weights and ‘Harris numbers’
THE SHEET DOESN’T GIVE THE WHOLE PICTURE, BUT AS A REFERENCE IT WORKS FOR ME
SHEET 2 Body modes and weights before and after edgework and set-up, along with a sketch of the C-bout archings
IT’S REMARKABLE HOW QUICKLY ONE FORGETS HOW A VIOLIN SOUNDED EVEN A MINUTE OR SO AFTER HEARING IT
I will quite often continue to adjust the top plate after finishing the bass-bar, and in the example shown I have added a veneer under the soundpost and some cleats above and below the f-holes which have brought all the numbers up a little. I rarely alter the backs after gluing them to the ribs. An interesting example of the advantages of putting numbers into a visual context can be seen between the free back and the back on the ribs. Every violin will basically follow this pattern where modes 2 and 5 move inside the parameters of the same modes in the free plate. On a cello, Mode 5 will invariably move higher than it was before, the result of proportionally higher and heavier ribs.
Ifinalise the edge overhang and do the purfling and edgework all after closing the body of the violin. Sheet 2 helps me track this process via body modes and weight. It would be better to be able to record relative amplitudes of these modes, but for now this will do. I put a soundpost in to begin with, and usually just record A0, the Helmholtz frequency, for the first few steps, adding modes B0, B1+ and B1- once the neck is in.
Finally, a simple tracing of the finished C-bout arches from a contour gauge are a useful reference, and complete my documentation of the instrument. The straight line from the edge tangent to the arch helps me to clarify the visual impression.
Sheet 3 is intended to document the sound of the violin, but I’ve found it just as useful as a guide when working with players on adjustment. I think it’s remarkable how quickly one forgets how a violin sounded even a minute or so after hearing it. This is particularly noticeable when adjusting a soundpost with a player or running listening tests with other violin makers. There have been many attempts to develop a language to describe tone to help overcome some of these difficulties. My approach is to create a simple memorable structure for the sound of a particular violin, upon which more complex observations can be placed. Similar techniques are used by memory champions who can create a ‘mind map’, usually based on a simple imagined structure or story, to which complex information can be attached.
The star graph deals with the tonality of the violin and I usually deal with this first. It’s made up of three scales, which cover what are to me the most obvious aspects, or parameters, of the vocal quality of the violin. The three scales are ‘Bright–Dark’, ‘Open–Covered’ and ‘Textured–Clean’. I make a mark on each scale to reflect its tone in that particular parameter. It’s important not to see these as positives or negatives; they are simply tonal attributes. Now I have made a fundamental judgement on the tone and have a memorable little pictogram in my mind. The example shows what would be a typical new violin for me. The mark is in the mid-point on the ‘Bright–Dark’ scale, somewhere in the direction of open on ‘Open–Covered’ and strongly in the direction of textured on ‘Textured–Clean’.
Once you have made your assessment on these three aspects, and have established the picture, you can easily add (and remember) things like ‘the E string is sweet’, ‘the A lacks focus above third position’ and so on without losing sight of where you were. With very little practice you can get very quick at making these three assessments, and add other observations even without the help of the sheet. You can even use the points of the graph to attach any ‘E string is sweet’ type of observations.
The ‘Response’ side is arranged as a parallel series of scales as these aspects seem less interconnected. Again I have tried to keep things as simple as they can possibly be. The third scale, on volume, is the only one limited by unwanted extremes. In my observation, makers as well as musicians are very quick to pick up on an instrument’s comparative volume, so putting this scale last means it doesn’t dominate the general impression.
MAKERS AS WELL AS MUSICIANS ARE VERY QUICK TO PICK UP ON THE COMPARATIVE VOLUME OF AN INSTRUMENT
SHEET 3 Asimple record of the instrument’s tone quality, along with a fingerboard diagram
Finally I have a diagram of the strings and fingerboard, which can be useful to mark the positions of wolfs, and tonal quirks specific to position on a string. This makes a great guide to assess problems and check what has been changed by any adjustment.
I am by no means an expert in acoustics. I have a very broad-brush approach to the business of acoustic testing, and prefer to focus on those other aspects of making that seem to have the biggest impact. I have genuinely found that recording data in a more visual way has helped me enormously in my day-to-day making, and I hope these simple sheets will provide other makers with some inspiration for their own charts.