Soundpost Material and Seasonal Changes

[MaestronetLurker]

I keep thinking about the fact that arching rises and lowers with the season, and we have to adjust posts because they don't change with the seasons.  Is anyone aware of experimentation with posts made in a way so they expand and contract somewhat with humidity changes? 

 

I know there's some debate over the material of the post, and whether it has a tonal effect.  If sound travels faster along the endgrain of spruce, then it's an obvious choice.  I've heard some well renowned people argue on either side of the question of whether more or less dense posts actually effect the tone.  Me experience suggests that the material does influence the sound (I recognize it may be my expectations guiding my perception), but the arguments to the contrary have been compelling. 

 

If material choice doesn't effect the function of the post (within reason), then I would think that a grain orientation that has the endgrain facing East-West rather than Up-Down would cause it to expand and contract similarly to the changes in arching.  Of course the most obvious structural issue with this orientation is the tendency to split the post, but I imagine a washi paper wrap would keep it secure.  The other issue might be a tendency for the post to flex thereby absorbing some of the energy that would otherwise transfer through the post.

 

In any case, I probably won't make it back to weigh in on this any time soon, but I thought some others on MN enjoy speculating on the subject.  I'd love to hear any real experience related to post material choices.

[Don Noon]

I haven't done any very careful tests on post material, but I have put in temporary posts of hardwood dowel, and made a post ~2mm diameter (with wider flared ends so as not to damage the plates), and haven't found any significant differences.  Perhaps if I was a player who was hyper-sensitive to things, the story might be different... but in any case I think the effects would be quite small.

 

I doubt that you'd have any success in trying to compensate for seasonal (humidity) distortions with a variable soundpost.  First, I don't think you'll find any material that could support the vertical forces and still move ~.5mm or so in length over the  fairly short post length.  Even if you could find such stuff, it would only push up the post side, not the bass bar.  And who knows what kind of other distortions, stresses, or damage it might be causing.

 

A better solution would be to determine exactly what is causing the seasonal variations in neck projection, and engineer a solution based on that.  I suspect most of the change is from the bi-material of the neck heel (tangential) vs. the neck block (longitudinal).  First thing I'd try would be a fat dowel or graphite rod through the heel of the neck, paired with a tangentially oriented neck block.

[~ Ben Conover]

Buy a load of good post material, and use it. 
End of. 

[FiddleDoug]

There is almost no longitudinal dimensional change in wood with changes in humidity.

"Wood is fairly stable along its longitudinal direction, parallel to the grain. Green lumber shrinks only 0.01 percent of its length as it dries. An 8-foot-long board will move only 3/32 inch."

And this is for GREEN lumber. Seasoned wood would have less change. For a 50 mm soundpost, 0.01% would be 0.005 mm. Remember, seasoned wood is less! Any changes are from the body/neck of the instrument, and I expect that each instrument would be different.

[Will L]

Violins change perceptibly with the slightest variations in the air. With many violins I have been able to feel/hear changes just by opening a window, for example.  It's hard to know for sure but I assume some violins are more sensitive than others; and no doubt there are different factors which would make tests very hard to do.

 

I think the best we can hope for is to have a violin which will still function well even though it is continually changing.  Maybe another way of putting it is that it will stay within a certain range of playability no matter how extreme the changes in the air.  I used to find some violins always remained within a good range, even when they weren't at an optimum, while others would become unplayable.  

 

When I used to be around the old master instruments I remember having at least the illusion that they were more stable than lesser and/or newer instruments.  Very subjective, I know.  Once, when I was having huge troubles with my personal violin, I began asking other violinists about their impressions. I asked a fellow named Paul Shure if he had any trouble keeping his instrument adjusted (it was some variety of Guarneri, not dG).  And he told me he hadn't  felt the need to have it adjusted for perhaps 20 years.  Well, either that violin was very stable and worked adequately under all sorts of conditions. Or the violinist was not very sensitive.  Or, of course, he might have been able to fight through discomfort better than I could.  

[David Burgess]

I think Maestronet Lurker was (in part) suggesting a change in grain orientation of the soundpost, such that it might better expand and contract in a more similar way to changes in distance between the top and the back, as moisture levels change.

[Bruce Carlson]

Violins change perceptibly with the slightest variations in the air. With many violins I have been able to feel/hear changes just by opening a window, for example.  It's hard to know for sure but I assume some violins are more sensitive than others; and no doubt there are different factors which would make tests very hard to do.

 

I think the best we can hope for is to have a violin which will still function well even though it is continually changing.  Maybe another way of putting it is that it will stay within a certain range of playability no matter how extreme the changes in the air.  I used to find some violins always remained within a good range, even when they weren't at an optimum, while others would become unplayable.  

 

When I used to be around the old master instruments I remember having at least the illusion that they were more stable than lesser and/or newer instruments.  Very subjective, I know.  Once, when I was having huge troubles with my personal violin, I began asking other violinists about their impressions. I asked a fellow named Paul Shure if he had any trouble keeping his instrument adjusted (it was some variety of Guarneri, not dG).  And he told me he hadn't  felt the need to have it adjusted for perhaps 20 years.  Well, either that violin was very stable and worked adequately under all sorts of conditions. Or the violinist was not very sensitive.  Or, of course, he might have been able to fight through discomfort better than I could.  

I met Paul Shure in Los Angeles while I was working at Weisshaars and got a chance to photograph his violin. Or perhaps, one of his violins, it was attributed to Michel Angelo Bergonzi with a beautiful one piece back.

 

Bruce

[go_oa]

I remember hearing about soundposts with the length established by a steel spring.  That would be the adjustable post.  Unfortunatly the post would also act as a filter, removing some frequencies from its transmissions to the back.  Somehow the spring loaded sounpost has not become popular!

[Will L]

I met Paul Shure in Los Angeles while I was working at Weisshaars and got a chance to photograph his violin. Or perhaps, one of his violins, it was attributed to Michel Angelo Bergonzi with a beautiful one piece back.

 

Bruce

I believe that may have been his wife's violin, since she was known to have a Bergonzi. I never saw it.  But the violin he showed me was a "lesser" Guarneri, according to him.  

 

There is an interesting obituary on him:

 

http://variety.com/2011/film/news/paul-shure-movie-score-concertmaster-dies-at-89-1118032505/ 

[PASEWICZ]

Lurker,

This is an interesting thought experiment, but as was mentioned earlier, there are so many things that change with RH, controlling the environment is the only practical answer.

[MaestronetLurker]

I doubt that you'd have any success in trying to compensate for seasonal (humidity) distortions with a variable soundpost.  First, I don't think you'll find any material that could support the vertical forces and still move ~.5mm or so in length over the  fairly short post length.  Even if you could find such stuff, it would only push up the post side, not the bass bar.  And who knows what kind of other distortions, stresses, or damage it might be causing.

 

A better solution would be to determine exactly what is causing the seasonal variations in neck projection, and engineer a solution based on that.  I suspect most of the change is from the bi-material of the neck heel (tangential) vs. the neck block (longitudinal).  First thing I'd try would be a fat dowel or graphite rod through the heel of the neck, paired with a tangentially oriented neck block.

 

According to this shrinkage calculator http://www.woodbin.com/calcs/shrinkulator/ a 55mm soundpost made of wood with average shrinkage rates (4%Radial, 8% Tangential) would shrink by 0.34mm radially or 0.7mm tangentially when going from 65%RH to 35%RH, so depending on how much you wanted the post to move you could select the grain orientation accordingly.  You could select from a wide range of materials and have similar rates of shrinkage.  If your post tension is not enough to dent the spruce, then it shouldn't be compressing a spruce post that was not oriented in the traditional direction.

 

As for what is changing, I'm working on putting together a study to answer that exact question.  I'm not a scientist or a researcher, but as the concept develops I'll reach out to some folks who know what's what.  Essentially what I would like to do is send musicians with data loggers to track the RH where the instrument is played, (possibly one that is built into an end button to read inside as well with the actual logger clamped chinrest style to a rib or disassembled and the circuitry attached under the chinrest?).  The musicians would periodically weigh the instruments and note tonal changes.  This should be able to be lined up with the data logger time stamps to sort out environmental changes.  Then I'd like to periodically measure string heights, neck extension, and height of the bass and treble f wings.  I'd like to CT scan the instruments near the low point of MC and high point of MC and compare what has changed with different instruments.  I'll need some funding to make that all happen, but it could be informative as to how the instrument changes as it moves from a humidity controlled case to an uncontrolled hall.  How quickly does it change?  Could this perhaps inform us as to playing the curve a little better when it comes to humidity changes in instruments?  Might all be for nothing, but I'd like to investigate it.  Initial data collection suggests some instruments change 0.1mm for every 1g of weight gain, and I've seen a few newish violins change by 16g from winter (35%-40%RH to summer 65%-70% RH).  I want get a larger sample size with more old instruments included. 

[MaestronetLurker]

I remember hearing about soundposts with the length established by a steel spring.  That would be the adjustable post.  Unfortunatly the post would also act as a filter, removing some frequencies from its transmissions to the back.  Somehow the spring loaded sounpost has not become popular!

 

I'm not sure the post does act as a filter.  I used to think it does, and I haven't been able to shake the idea in spite of some compelling arguments to the contrary.  Really, though, I think the post wouldn't filter any low frequencies, and the back isn't really radiating upper frequencies, nor do they come from inside (as far as I understand).  The post pumps the back and supports the top, but I think it's by direct force and the frequencies traveling through it probably don't matter.  Then again, I could be completely wrong.  I'd like to hear about people's experiments and observations in this regard.

[PASEWICZ]

M.L.,

The effects on an instrument in varying RH are pretty well known, what do you hope to gain that is not already common knowledge?

[MaestronetLurker]

Lurker,

This is an interesting thought experiment, but as was mentioned earlier, there are so many things that change with RH, controlling the environment is the only practical answer.

 

I think this is true, but controlling the concert spaces doesn't seem to actually be practical.  I don't get the impression that concert halls are usually tightly controlled in terms of humidity.  Even so, there are many spaces including outdoor spaces where things get a little crazy.  I suspect that the two biggest effects on the seasonal tonal change are the post compression and damping from the wood absorbing moisture.  Not much to be done about the damping, but if a post could adjust automatically in a similar way to the instrument that would be pretty interesting.

[MaestronetLurker]

M.L.,

The effects on an instrument in varying RH are pretty well known, what do you hope to gain that is not already common knowledge?

 

I think it's hard to predict entirely what may be learned from studying the effect of humidity changes in greater depth.  One thing that comes to mind is the exact way that the top and back arches change with the seasons.  By looking at a CT scan of numerous instruments you could get sort of an average sense of how it happens, and how much various areas change.  Some instruments seem to be more stable, and if some instruments in the study are not perceived to have significant seasonal performance issues, then it may be interesting to look at what makes them different.  Especially if they are changing as much as the others, but aren't having issues.  On the contrary, if some instruments are more stable, it may be interesting to investigate why that might be.

 

One thing that I'm curious about is consistency of shrinkage rates and how a different shrinkage rate of the top vs. the back may change things.  We generally consider longitudinal shrinkage to be nil, but over a long area it has an effect.  You don't tend to see longitudinal shrinkage data, but I was taught that the volumetric shrinkage is the overall rate, and if you want to see longitudinal shrinkage you add radial and tangential shrinkage and deduct it from volumetric shrinkage.  That would give me a longitudinal shrinkage rate of 1.1% for Engelmann, 0.8% for Norway Spruce, 0.2% for Red Spruce.  Can't find data for Acer Platanoides, but Big leaf is 0.8% and Acer Pseudoplatanus (the other European Maple) has virtually 0% longitudinal shrinkage.  I don't know how consistent the shrinkage rate is within a species or how accurate these numbers are, but if you imagine how a top with 1% longitudinal shrinkage and back with significantly less longitudinal shrinkage would change seasonally the effect on neck extension might be significant and perhaps a top and back that move more similarly to eachother would have less perceived seasonal change.  If it turned out that greater stability could be had by matching woods that will move at more similar rates, then that would be interesting. 

 

Over the length of a cello back a shrinkage rate of 1% translates to over 1mm of shrinkage from 65%RH to 40%RH.  If there is a difference of 1/2mm in the way the top and back move longitudinally it would make a significant difference in neck angle.  Anyways, my overall point is just that there may be a difference between what we think is happening and what is actually happening, and I'd like to investigate that a bit more closely.  Hopefully it yields something worthwhile for the effort.

[Jim Bress]

ML, I've used HOBO U23 data loggers for temp/RH quite a bit.  You can get them standard with six foot cables and sensors as small as 5 mm in diameter which would allow you to insert the sensor through the bass f-hole.  Longer cables may be available (not sure).  I recorded every hour 12 hours (ecosystem stuff) over a 2 year period outdoors.   The sensors were very reliable for me.  You can set up the recording interval as often as every second.  I think internal temperature/RH changes during playing in different conditions  (practice room, studio, concert hall, outdoor stage, etc.) could be interesting.

 

-Jim 

[PASEWICZ]

I think it's hard to predict entirely what may be learned from studying the effect of humidity changes in greater depth.  One thing that comes to mind is the exact way that the top and back arches change with the seasons.  By looking at a CT scan of numerous instruments you could get sort of an average sense of how it happens, and how much various areas change.  Some instruments seem to be more stable, and if some instruments in the study are not perceived to have significant seasonal performance issues, then it may be interesting to look at what makes them different.  Especially if they are changing as much as the others, but aren't having issues.  On the contrary, if some instruments are more stable, it may be interesting to investigate why that might be.

 

One thing that I'm curious about is consistency of shrinkage rates and how a different shrinkage rate of the top vs. the back may change things.  We generally consider longitudinal shrinkage to be nil, but over a long area it has an effect.  You don't tend to see longitudinal shrinkage data, but I was taught that the volumetric shrinkage is the overall rate, and if you want to see longitudinal shrinkage you add radial and tangential shrinkage and deduct it from volumetric shrinkage.  That would give me a longitudinal shrinkage rate of 1.1% for Engelmann, 0.8% for Norway Spruce, 0.2% for Red Spruce.  Can't find data for Acer Platanoides, but Big leaf is 0.8% and Acer Pseudoplatanus (the other European Maple) has virtually 0% longitudinal shrinkage.  I don't know how consistent the shrinkage rate is within a species or how accurate these numbers are, but if you imagine how a top with 1% longitudinal shrinkage and back with significantly less longitudinal shrinkage would change seasonally the effect on neck extension might be significant and perhaps a top and back that move more similarly to eachother would have less perceived seasonal change.  If it turned out that greater stability could be had by matching woods that will move at more similar rates, then that would be interesting. 

 

Over the length of a cello back a shrinkage rate of 1% translates to over 1mm of shrinkage from 65%RH to 40%RH.  If there is a difference of 1/2mm in the way the top and back move longitudinally it would make a significant difference in neck angle.  Anyways, my overall point is just that there may be a difference between what we think is happening and what is actually happening, and I'd like to investigate that a bit more closely.  Hopefully it yields something worthwhile for the effort.

Sounds good.

[Don Noon]

According to this shrinkage calculator http://www.woodbin.com/calcs/shrinkulator/ a 55mm soundpost made of wood with average shrinkage rates (4%Radial, 8% Tangential) would shrink by 0.34mm radially or 0.7mm tangentially when going from 65%RH to 35%RH, so depending on how much you wanted the post to move you could select the grain orientation accordingly.  You could select from a wide range of materials and have similar rates of shrinkage.  If your post tension is not enough to dent the spruce, then it shouldn't be compressing a spruce post that was not oriented in the traditional direction.

 

The shrinkage/expansion values are for unconstrained pieces of wood, and accounting for the stresses in a soundpost application will likely give some other result.  This is a very complex indeterminate structure, where you need to look at the relative stiffness of the body vs. the soundpost to see where things end up.  For a normal soundpost, the stiffness is extremely high, and will undoubtedly dominate the stiffness of the body.  For a tangential post, softer by a factor of 20, it isn't necessarily obvious.  Rubber bands have a huge length change with temperature... but you can't push things around with rubber bands.

 

I'm not sure the post does act as a filter.  I used to think it does, and I haven't been able to shake the idea in spite of some compelling arguments to the contrary.  Really, though, I think the post wouldn't filter any low frequencies, and the back isn't really radiating upper frequencies, nor do they come from inside (as far as I understand).  The post pumps the back and supports the top, but I think it's by direct force and the frequencies traveling through it probably don't matter.  Then again, I could be completely wrong.  I'd like to hear about people's experiments and observations in this regard.

 

Nothing is infinitely stiff, and therefore everything is a spring if you look closely enough.  The normal soundpost is a fairly stiff spring, with some masses (the plates) at each end.  There will be some frequency where the mass/spring system is in resonance.  I made some very rough guestimates and arrived at ~5000 Hz as the lowest frequency where the post would cause resonance.  I don't know how accurate the Strad3D animations are, but there appears to be top/back movements out of phase in the soundpost area at somewhat lower frequencies than that, implying that the soundpost is flexing axially to a significant degree.

 

A tangential (or otherwise less stiff) post would reduce the resonant frequencies significantly, unless you were to increase the area proportionally (like by a factor of 20).  I doubt the low frequencies would be much affected; the main area where you'd see effects would be the highs.

[MaestronetLurker]

The shrinkage/expansion values are for unconstrained pieces of wood, and accounting for the stresses in a soundpost application will likely give some other result.  This is a very complex indeterminate structure, where you need to look at the relative stiffness of the body vs. the soundpost to see where things end up.  For a normal soundpost, the stiffness is extremely high, and will undoubtedly dominate the stiffness of the body.  For a tangential post, softer by a factor of 20, it isn't necessarily obvious.  Rubber bands have a huge length change with temperature... but you can't push things around with rubber bands.

 

 

Nothing is infinitely stiff, and therefore everything is a spring if you look closely enough.  The normal soundpost is a fairly stiff spring, with some masses (the plates) at each end.  There will be some frequency where the mass/spring system is in resonance.  I made some very rough guestimates and arrived at ~5000 Hz as the lowest frequency where the post would cause resonance.  I don't know how accurate the Strad3D animations are, but there appears to be top/back movements out of phase in the soundpost area at somewhat lower frequencies than that, implying that the soundpost is flexing axially to a significant degree.

 

A tangential (or otherwise less stiff) post would reduce the resonant frequencies significantly, unless you were to increase the area proportionally (like by a factor of 20).  I doubt the low frequencies would be much affected; the main area where you'd see effects would be the highs.

 

 

I don't know how the shrinkage and expansion would be changed by being constrained, but I wouldn't think the compression on the post is so much that it would counteract the expansion with increased MC.  

 

I wouldn't be surprised if the flexibility of the post oriented cross-grain was a problem, but I think it could be resolved.  I would think that flexibility of the post would have to be increased quite a bit before it were a noticeably factor, but that's a wild guess on my part.  I don't see a practical way of increasing the area by a factor of 20, but I can imagine a post that is a wider diameter tapered down to normal diameter just where it meets the plates. As long as it can be inserted through the lower lobe, then it should work.  

 

Another option could be a post set within a carbon fiber tube slotted on one side.  The post would not be glued into the tube, just pressed in so it clamps onto the post while allowing it to move vertically.  Should add stiffness, but allow vertical expansion.  At least those are the most obvious attempts at a solution that come to my mind. 

[Don Noon]

Even if this post idea works, there's still the problem that it only works on the treble side, and does nothing for the bass bar.  And if the top plate torsional stiffness is more to the center than at the bass bar, then the post adjustment will make the bass side go the other way, and make things worse.

 

I still believe that most of the projection variations originate at the neck/body joint, and would be most productive to solve there.

[David Burgess]

Even if this post idea works, there's still the problem that it only works on the treble side, and does nothing for the bass bar.  And if the top plate torsional stiffness is more to the center than at the bass bar, then the post adjustment will make the bass side go the other way, and make things worse.

 

I still believe that most of the projection variations originate at the neck/body joint, and would be most productive to solve there.

Don, I don't think he's talking so much about neck projection, and I generally wouldn't try to change that with a soundpost anyway. Those of us who live in climates with extreme seasonal variations notice that the arching height can change enough, that sometimes a different length post will be required for different seasons. More frequently, the tightness or looseness of the soundpost will change enough to alter the sound.

 

Not that the two are totally unrelated. When the top expands from increased moisture content, it will increase the arching height, contributing to a decrease in the measured fingerboard height. There are several other factors involved. Even the very slight increase in the length of the back will lower the fingerboard a little bit.

[Don Noon]

Yep, I re-read the OP, and it doesn't seem to be an attempt to compensate for projection.  I just made that erroneous assumption.  Although I have noticed tonal changes with seasonal variations, I think it is mostly due to the plate acoustics and moisture content in the wood, not the post fit.  I haven't been around any picky violinists that can detect a need for season-specific posts. 

[MaestronetLurker]

I think that the length of post is the biggest factor in seasonal performance/tonal changes.  Sure there's some damping from more moisture in the plates, and other factors will contribute, but elite level players will often notice issues that can be dealt with by adjusting the post.  A friend of mine who worked at a high-end shop in MN told me that most of the Minnesota Orchestra (by his recollection) had seasonal posts and bridges.  Even on violins.  That isn't the norm around here, but it makes sense to me.  Particularly if you have a record of the seasonal shift of the instrument through measurements of neck projection, string heights, and weight (and perhaps height of the upper fhole wings?) to show how and when it is changing over the seasons.  Pairing this information with the players observations about tone and performance I think it would be possible to play the curve a little better, so there's more overlap between the two seasonal setups so they don't have to suffer through a period where it doesn't perform well.

[MaestronetLurker]

ML, I've used HOBO U23 data loggers for temp/RH quite a bit.  You can get them standard with six foot cables and sensors as small as 5 mm in diameter which would allow you to insert the sensor through the bass f-hole.  Longer cables may be available (not sure).  I recorded every hour 12 hours (ecosystem stuff) over a 2 year period outdoors.   The sensors were very reliable for me.  You can set up the recording interval as often as every second.  I think internal temperature/RH changes during playing in different conditions  (practice room, studio, concert hall, outdoor stage, etc.) could be interesting.

 

-Jim 

 

Thanks for that.  This might work for building an endbutton around the cable and leaving the sensor inside the instrument, then mounting the other portion of the device externally.  Alternatively, I also considered something like this with a removable external portion that would have a plug with a magnetic catch.  Something like the power cable attachment of an Apple laptop that unplugs without damage if it gets yanked.  My thought would be that a musician could have the device in a coat pocket or somewhere and have it plugged to the endpin, and if it gets yanked somehow it wouldn't pull the instrument too hard.  It could also be removed if there were a performance where the musician didn't want to have this extra component attached.  Not sure whether it's practical to modify a device to have a cable that releases easily like that, but it has been on my mind. 

[David Burgess]

 

 

As for what is changing, I'm working on putting together a study to answer that exact question.  I'm not a scientist or a researcher, but as the concept develops I'll reach out to some folks who know what's what.  Essentially what I would like to do is send musicians with data loggers to track the RH where the instrument is played, (possibly one that is built into an end button to read inside as well with the actual logger clamped chinrest style to a rib or disassembled and the circuitry attached under the chinrest?).  The musicians would periodically weigh the instruments and note tonal changes.  This should be able to be lined up with the data logger time stamps to sort out environmental changes.  Then I'd like to periodically measure string heights, neck extension, and height of the bass and treble f wings.  I'd like to CT scan the instruments near the low point of MC and high point of MC and compare what has changed with different instruments.

ML, Bruce Carlson has already collected a lot of data like this on the Cannone, and I think it's been published. Maybe you could contact him to find out how to access it.