Melodeon.net Forums

Discussions => Instrument Design, Construction and Repair => Topic started by: Stiamh on September 29, 2011, 03:16:01 PM

Title: Soundboards again
Post by: Stiamh on September 29, 2011, 03:16:01 PM
The prevailing opinion seems to be that "that board" is not involved in and has no effect on sound production. 

Has this idea ever been properly tested?

It reminds me of the argument of certain physicists that the material a flute is made of has no bearing on the tone of the instrument, this being purely a product of the dimensions of the bore and other factors unconnected with the material.

I am no expert as a flute player but I don't buy this argument and I'm sure that many expert flute players don't buy it either. I have played barrowloads of whistles though and I think that plastic, aluminium, and wooden varieties each tend to have a characteristic timbre. 

This makes me view the prevailing opinion with a small degree of scepticism. Has anybody fitted the same instrument with soundboards of different materials and tested the effect of aluminium, plywood, solid wood, etc.?
Title: Re: Soundboards again
Post by: mory on September 29, 2011, 06:08:14 PM
Is it a prevailing opinion or just the one being voiced?
Wood and alloy sound boards seem to have such powerfully different characteristics I personally feel that the material is involved, if there is no vibration of the sound board presumably this can be measured, if there is vibration I think it would be hard to deny an involvement in sound production, secondary resonance or otherwise. My small experiments for my own interest have at one time or another involved replacing a couple of sound boards with ones other than the original material and found them to have completely different characteristics, Some one also gave me a poker work made of 8mm perspex apart from the reed blocks bellows and mech, it was a less than satisfying result, I felt, as it's sound was nothing like the rich full sound of the original and seemed very harsh and lacking in body/depth or any real melodeon/button accordion warmth and character, so much so that apart from robbing the reeds n mech etc, it went into the trash. I'm sure I will be informed otherwise soon enough, but that's my experience. Isn't it wonderful to have so many experts in one place  ;D mory
Title: Re: Soundboards again
Post by: Pete Dunk on October 02, 2011, 12:17:38 AM
I'd believe in a theory involving sound reflection to do with the surface density of the materials involved. A 'soundboard' as such doesn't make much sense to me at all because it is small, bound on all sides and has massive, rigid reed blocks mounted at right angles to the smallest dimension. The solid body of an electric guitar has an impact on it's sustain but not on the acoustic properties. I'm not saying that the use of alloys or varnished/dense woods inside accordions will have no effect on the overall sound but it will be more about bouncing it around efficiently and getting it out quickly than some mystical vibration of one material versus another.

Coat!
Title: Re: Soundboards again
Post by: blafleur on October 02, 2011, 12:32:49 AM
On 2 accordions I just finished and had put 2 layers of felt on the valves (pallets) with the leather (as opposed to my usual one layer felt), hoping to make it seal better and be quieter.  Both lacked the usual punch compared to those I've finished before, and after seeing a comment by Peter Hyde, I removed one layer, and now much better sound.  So if one little layer of felt, outside of but against the soundboardfaceplatefondo can make that much difference, it sure stands to reason that the make up of the soundboardfaceplatefondo would make a sound difference also.  I am building one now that will have solid spruce core with very thin layer of veneer on both sides, as opposed to my usual aircraft plywood, anxious to see if there is a tonal difference, and will report on it.

But, just more slightly tested opinions........
Title: Re: Soundboards again
Post by: Andrew Wigglesworth on October 02, 2011, 12:38:33 AM
On a violin there is a mechanical link between the string and the soundboard. Take away the mechanical link and the note from the string does not get amplified through the resonances built up in the instrument.

Play an electric violin unplugged and see how much sound you get wthout a soundboard.

In an accordion the mechanical link between the reed and the rest of the instrument plays no part in the creation of the note. If it did than you'd still hear it playing after you lifted your finger off the button and closed the pallet. The reed is still vibrating, but there is no note being produced, not even a vestige of one.

The vibrations in the body of an accordion are being created by air waves. This is the transmission of sound (and it's reflection etc), not it's creation.

To call part of a melodeon a "sound board" would be like calling the lid of a cardboard box (and no other part of it) a soundboard because you had a radio playing inside it.
Title: Re: Soundboards again
Post by: Owen Woods on October 02, 2011, 01:10:11 AM
All true. But I can't believe that the soundfacefondoplateboard doesn't vibrate at all. I think it reasonable to assume that it does vibrate, which will change the sound generated by the instrument, due to the loss of energy. So different materials, with different flexibility will give different sounds. We can see this from anecdotal evidence.

Another question, arguably more important, is whether this is a good thing. I know that many are of the opinion that the more rigid it is the better, and that the ideal box would be totally rigid and reflecting. As I've said before, I'm not sure that this would be the case, as the sound of a pure free reed siren sound is not very pleasant, rather harsh and piercing. It may well be that absorption by the components of the box and sympathetic vibration actually improve the sound. However, I don't have any evidence for this and I don't think anyone has done a proper study of it. I will have access to equipment that could tell me this year... but I'm supposed to be using it for other things :P
Title: Re: Soundboards again
Post by: blafleur on October 02, 2011, 04:11:22 AM


In an accordion the mechanical link between the reed and the rest of the instrument plays no part in the creation of the note. 


I don't think I can agree that it plays no part, it may not play a big part, but it has to play a small part in the quality of the sound, however small.  Even if minute, it's worth exploring because you put a few minute differences together and it add up to a noticeable difference.
Title: Re: Soundboards again
Post by: IanD on October 03, 2011, 04:06:25 PM


In an accordion the mechanical link between the reed and the rest of the instrument plays no part in the creation of the note. 


I don't think I can agree that it plays no part, it may not play a big part, but it has to play a small part in the quality of the sound, however small.  Even if minute, it's worth exploring because you put a few minute differences together and it add up to a noticeable difference.

If you go down this route -- everything however small can make a difference -- you'd better have a very large development budget because you're sure going to have to try out a lot of different combinations.

The correct engineering approach is to start with the things that we know make a big difference and can explain why this is so and how they might be improved (reeds, reed tuning/setting reed block volume/shape/bracing, pallet hole size/thickness, grille...) and get those right first. Then if you want you can look at the things which engineering says almost certainly don't make a difference but might conceivably make a small one like pallet board material.

Sorry to keep using the word "engineering" but it's the science of solving real-world problems using analytical method rather than mumbo-jumbo, it's what brings you things that work instead of homeopathy and crystals ;-)

Ian

P.S. Yes I know Stradivarius knew nothing about engineering, but we can still use it to explain why his instruments sound so good. Except that in a double-blind trial with discriminating listeners against a modern violin developed scientifially the Strad came second...
Title: Re: Soundboards again
Post by: Stiamh on October 03, 2011, 04:25:17 PM
The debate illustrates why I asked the question in my OP - has the effect of soundboard material ever been properly tested?

I don't think you would need a very large development budget to do this.

In fact there was an answer to the question in another thread (in the 21st century box thread I think it was) from someone who had tried both wooden and metal soundboards in the same model of box and found there was a distinct difference.



Title: Re: Soundboards again
Post by: blafleur on October 03, 2011, 05:55:32 PM
"properly"?  That doesn't sound as much fun as tripping and stumbling around and groping for opinionated guesses. ;D

Problem with doing much scientifically is a lot of what we'er looking for amounts to personal preference.

Title: Re: Soundboards again
Post by: IanD on October 03, 2011, 06:18:35 PM
The debate illustrates why I asked the question in my OP - has the effect of soundboard material ever been properly tested?

I don't think you would need a very large development budget to do this.

In fact there was an answer to the question in another thread (in the 21st century box thread I think it was) from someone who had tried both wooden and metal soundboards in the same model of box and found there was a distinct difference.


You'd need a large budget because if you start looking at the unlikely things (of which there are many) as opposed to the likely ones (of which there are few) you have an awful lot to try out. How about the pallet material (wood? plastic? metal?). How about the exact shape of the pallets? How about round vs. square pallet holes (with the same area)? How about [insert unlikely factor here]?

Was the pallet board trial done with materials of the same thickness, was it repeated with the reed blocks swapped between the two instruments, and did the player or the listener know which one was being played?

If not you haven't shown there's a difference due to the pallet board material, not the reeds (even if nominally identical) and observer bias -- either by the player or the listener.

Sorry to go on about this, but just the observer bias caused by knowing which one is which has been the downfall of many claimed differences which have been shown to be untrue under properly controlled conditions (double blind or ABX testing).

For example, proponents of sound differences between exotic (and expensive) audio cables have been consistently unable under these conditions -- actually, every single time -- to show they can hear any difference that isn't simply due to basic electrical properties (LRC) -- they can only tell which is which ("this sounds so smooth") if they already know the answer. This doesn't stop them continuing to insist to all and sundry that a $1000 cable will make their hi-fi sound better.

(and of course if the listener believes it does, then it does to them -- the placebo effect again)

The simple fact is that free reed instruments are completely different to stringed or beating reed/pipe instruments because the tone generation is almost entirely determined by the free reed and the acoustic path to the outside air, and the acoustic impedance seen by the reed. Little energy is coupled into the structure or radiates from or through it, unlike the other instruments (though resonances causing reed coupling can still affect the sound in specific circumstances, hence the strap between reed blocks). Any similarities drawn between free reed and other instruments regarding materials or structure are very likely to be invalid, they're a very different beast.

Those who disagree with this view, feel free to go and read the various books, papers and theses on the acoustics of musical instruments before drawing any conclusions -- also texts about experimentation, observer bias and how to remove it.

If (or when) somebody does the comparison properly and can demonstrate there's still a statistically significant difference, then I'll quite happily believe them. Until then I'll invoke Occam's Razor...

Cheers

Ian

P.S. Sorry if this comes over aggressively, but I see too many snake oil salesman to have any sympathy unless they can prove what they claim -- and music and musical instruments are more guilty of this than many other fields ;-)
Title: Re: Soundboards again
Post by: IanD on October 03, 2011, 06:23:49 PM
"properly"?  That doesn't sound as much fun as tripping and stumbling around and groping for opinionated guesses. ;D

Problem with doing much scientifically is a lot of what we'er looking for amounts to personal preference.



The problem is not personal preference (liking one sound more than another, perfectly valid), it's claiming there's a difference (based on opinion) when none has been shown to exist :-)

This has been shown to be very common in the audio world even if the observers don't realise they're doing it, musical instrument are exactly the same.

Science is largely about showing that a claimed result is not due to the prejudices -- unconscious or not -- of the experimenter or observer :-)

Ian
Title: Re: Soundboards again
Post by: oggiesnr on October 03, 2011, 08:00:34 PM
As I see it the basic requirements for a fair experiment would be ...


Not ruinously expensive but you'd certainly be able to buy at least one Briggs (or similar) for the cost  (:)

Steve
Title: Re: Soundboards again
Post by: IanD on October 03, 2011, 09:44:29 PM
As I see it the basic requirements for a fair experiment would be ...

  • a melodeon, probably single voice because that reduces the variables and is probably easier to measure (beats per second etc would mess up the analysis)

    a set of "soundboards" milled with openings of the same dimensions (a CNC job)

    a mechanical squeezing and button pushing device,  can't be a human as all pressures etc have to be identical

    measuring device to ensure same pressure across the reeds

    recording and analysing equipment.

Not ruinously expensive but you'd certainly be able to buy at least one Briggs (or similar) for the cost  (:)

Steve

Excellent idea. But maybe a 2 reed would be better, because one of the ways that any stiffness difference due to material might show up would be the "reed locking" that the strap between the blocks tries to prevent.

To put some numbers in, I reckon the reed blocks are at least 10x as stiff as a thin pallet board (wood or aluminium) because they're so much deeper (stiffness = thickness^2), which means ~90% of the stiffness comes from the reed blocks and ~10% from the pallet board. And thinking of the possibility of sound passing through the exposed pallet board from inside the box, a quick comparison to loudspeaker diaphragms of the same thickness and area with no suspension also suggests that this will be negligible too.

I await the results with interest... :-)

Ian
Title: Re: Soundboards again
Post by: Pete Dunk on October 03, 2011, 11:42:56 PM
We're getting down to the nitty-gritty of the subject here. Many of the theories expounded in this thread are based on hearsay and subliminal suggestion. I wouldn't know how to go about carrying out the kind of survey IanD is talking about but I do know that the method suggested sounds logical and would appear to remove any element of preconceived bias.

I'd happily undertake a blindfold test based on those principles.

Pete
Title: Re: Soundboards again
Post by: Owen Woods on October 04, 2011, 01:33:50 AM
Those who disagree with this view, feel free to go and read the various books, papers and theses on the acoustics of musical instruments before drawing any conclusions -- also texts about experimentation, observer bias and how to remove it.

I am interested, do you know whether there are any books or articles specifically about box acoustics?

My current thoughts are that the important factors are reed, block orientation (i.e. flat or upright) and chamber dimensions. These define the sound.

Bellows cross section will make a difference to the playability of the instrument, as will keyboard, mechanisms, pallets, etc. These shouldn't affect the sound though, or at least not significantly.

The casing of the instrument will affect the sound coming out, both in terms of projection and tone. The pallet hole area will affect how the reed sounds, as will the reed set.

Vibration of the reed blocks is likely to be unpleasant, as is vibration of the fondo (I suspect). For both, they should be as stiff as possible. The amount of space of the fondo covered by blocks may also have an effect on the vibration of the board. Fondo thickness might have an effect in that the hole to the block would be deeper. Likely to be negligible.
Title: Re: Soundboards again
Post by: Telemorris on October 04, 2011, 02:20:21 AM
I'm willing to start working on some of this AFTER I get my first melodeon built. Reading this thread has gotten me to thinking about how to deal with some issues to make sure the testing is for one variable and one variable only.

Since so much has been said about palletboards, I'd like to propose the following setup for testing different types of palletboard:

1. Create a set of reedblocks with reeds that move from one test instrument to another. No you can't do side by side comparisons, but the reeds, set of the tongue, chamber proportions, reedblock material would all stay the same. Make recordings and compare the recordings.

2. Create an instrument case such that palletboards can be swapped out, the reedblocks added and then play the instrument. I've seen some designs where this is fairly easy - a matter of a hand full of screws that hold the palletboard in place.

3. The palletboards all need to be treated exactly the same - same size and location of pallet holes, same dimension/area of pallet hole, same size of hold down blocks, etc.

4 What types/composition of palletboard should be tested? I'm thinking solid wood (this could also be divided into hardwood and softwood, for instance), plywood (aircraft ply), aluminium. Are there any other types that should be tested?

Ed
Title: Re: Soundboards again
Post by: oggiesnr on October 04, 2011, 07:15:29 AM
On further consideration, if the object is to discover whether the soundboard makes a difference (rather than which is the "best" soundboard material) then only two soundboards may need to be tested.  The trick would be to chose the two most likely "extreme" possibilities, say aluminium and cheap plywood of the same dimensions (tricky, do you define that by weight, thickness or flexilbility?), and test those under lab conditions.  If there is a difference then case is probably proved (at least for these purposes), if there is no difference then more testing would be required to check the general hypothesis that the material makes no difference.

Of course one might want to repeat the experiment with two different thickness soundboards of the same material to see if that has an effect.

Steve
Title: Re: Soundboards again
Post by: Andrew Wigglesworth on October 04, 2011, 08:22:19 AM
 8)
P.S. Sorry if this comes over aggressively, but I see too many snake oil salesman to have any sympathy unless they can prove what they claim -- and music and musical instruments are more guilty of this than many other fields ;-)

I'm just glad that you're here to make the points in a far better way than I could   8)
Title: Re: Soundboards again
Post by: Chris Ryall on October 04, 2011, 08:54:06 AM
This is 'at it's core' a materials science test and needs the sort of kit you'd find in a large engineering department .. with some sort of keen final year engineer to do the hard work  :|glug
Title: Re: Soundboards again
Post by: IanD on October 04, 2011, 11:38:38 AM
This is 'at it's core' a materials science test and needs the sort of kit you'd find in a large engineering department .. with some sort of keen final year engineer to do the hard work  :|glug

So maybe an interested engineering or RCM student (if there is such a thing) could do this as a final year project?

But I have to point out that though this would settle the argument, it would be much more useful to the box-making community to do some proper engineering research first on the things we think make a big difference, like reed material/profile/set, reed block chambers and bracing, pallet holes...

Answering the question about published material, there are a few papers about specific aspects of free-reed instruments and some information in books on musical acoustics, but unsurprisingly very little research has been done compared to stringed instruments like violin.
Title: Re: Soundboards again
Post by: Pete Dunk on October 04, 2011, 12:33:58 PM
"There is no problem, no matter how complex, which if looked at in the right way cannot be made even more complex."

Prof Bob Lowe, Leeds Metropolitan University.
Title: Re: Soundboards again
Post by: Ebor_fiddler on October 04, 2011, 05:21:40 PM
"When in trouble, fear or doubt,
 Run in circles, scream and shout!"

The late Robert Heinlein, God be good to him.
Title: Re: Soundboards again
Post by: Andy in Vermont on October 04, 2011, 06:34:12 PM
Sorry to keep using the word "engineering" but it's the science of solving real-world problems using analytical method rather than mumbo-jumbo, it's what brings you things that work instead of homeopathy and crystals ;-)

It is especially good at bringing things with massive, externalized costs, like industrial agriculture. And especially good at finding new ways to cause cancer.

I understand what Bryan is saying -- there are lots of little factors. No need for "engineering," just make two instruments that are otherwise as close as possible, and change one variable at a time, choosing whichever factor you'd like. If everyone making accordions followed Ian's methods, all we'd have is a bunch of blueprints and no way to make music.
Title: Re: Soundboards again
Post by: Owen Woods on October 04, 2011, 07:06:02 PM
On further consideration, if the object is to discover whether the soundboard makes a difference (rather than which is the "best" soundboard material) then only two soundboards may need to be tested.  The trick would be to chose the two most likely "extreme" possibilities, say aluminium and cheap plywood of the same dimensions (tricky, do you define that by weight, thickness or flexilbility?), and test those under lab conditions.  If there is a difference then case is probably proved (at least for these purposes), if there is no difference then more testing would be required to check the general hypothesis that the material makes no difference.

Of course one might want to repeat the experiment with two different thickness soundboards of the same material to see if that has an effect.

Steve

Well in a trivial sense, the soundboard will make a difference. If you have one out of aluminium and one out of rubber you'd be able to tell which was which! I sort of think that it's more useful to take the sort of materials that you might get (which would be solid wood, one or two types of ply and aluminium) and see whether there is a difference.
Title: Re: Soundboards again
Post by: Stiamh on October 04, 2011, 08:27:09 PM
Sorry to keep using the word "engineering" but it's the science of solving real-world problems using analytical method rather than mumbo-jumbo, it's what brings you things that work instead of homeopathy and crystals ;-)

It is especially good at bringing things with massive, externalized costs, like industrial agriculture. And especially good at finding new ways to cause cancer.

LOL. There's no thumbs up emoticon here, so...  :|glug
Title: Re: Soundboards again
Post by: Andy in Vermont on October 04, 2011, 08:45:18 PM
LOL. There's no thumbs up emoticon here, so...  :|glug

Funny you should mention that --  :|glug -- because wine is a good example of a product where the "mumbo jumbo" has faired very well, as of late. (And as for "double-blind" tests -- did you know that it has been shown that most wine drinkers cannot distinguish between red and white wine in blind tests of wines at room temperature!?!?)
One of the issues is that, to a certain extent, the results are beyond the control of the makers.
Title: Re: Soundboards again
Post by: IanD on October 04, 2011, 11:02:52 PM
Sorry to keep using the word "engineering" but it's the science of solving real-world problems using analytical method rather than mumbo-jumbo, it's what brings you things that work instead of homeopathy and crystals ;-)

It is especially good at bringing things with massive, externalized costs, like industrial agriculture. And especially good at finding new ways to cause cancer.

I understand what Bryan is saying -- there are lots of little factors. No need for "engineering," just make two instruments that are otherwise as close as possible, and change one variable at a time, choosing whichever factor you'd like. If everyone making accordions followed Ian's methods, all we'd have is a bunch of blueprints and no way to make music.

I think you're deliberately misunderstanding what I said, as well as distorting the meaning of the word "engineering" -- as an engineer, I find it sad that people have views like this. Without engineering you wouldn't have your car, bridges, ships, radio and TV (broadcast, cable and satellite), the internet, mobile phones, GPS, CAT/MRI scanners, aeroplanes -- the list is endless. Of course you can also find things not obviously to the benefit of humanity like the atomic bomb, but that's usually down to what politicians decide rather than the engineering profession.

Changing one variable at a time and seeing whether it makes a difference *is* science and engineering, and is exactly what I was suggesting -- but the results are only valid if you do the tests in a way that avoids observer bias. If you allow people to know what they're listening to (or seeing, or tasting) the results are almost guaranteed to be biased, as has been proved on innumerable occasions. Put simply, if you can't tell the difference without foreknowledge, there *isn't* a difference no matter how much you'd like there to be one and how many people believe there is one.

A perfect example is the modern Strad project I referred to earlier. It's a view very widely help among musicians, luthiers and distinguished listeners that violins improve with age due to some "playing-in" process which affects the sound, so a new instrument needs to be played a lot (months? years?) before it improves in tone. So they made two identical instruments (using mirror-cut wood from the same planks and so on), and tested them (properly!) after building to ensure that they sounded the same. One was then played for hours daily by working musicians for both practices and in concerts, the other was left unplayed in a carefully controlled environment. On several occasions over several years the sister violins were brought back together and tested again, and in no case was there any detectable difference in sound.

In other words, it doesn't matter how many people genuinely believe that the material a melodeon pallet-board is made from has a big effect on the sound, until *proven* it ain't so. It doesn't even matter if many of the people are respected makers or players of the instruments, because on so many occasions the "experts" have been shown to be wrong just like Joe Public.

Of course comparing a diamond pallet board with a rubber one would show that if you take anything to extreme it will have an effect, but here we're talking about the differences between "normal" reasonably rigid materials like different woods or aluminium.
Title: Re: Soundboards again
Post by: Stiamh on October 05, 2011, 02:21:14 AM
Ian, I don't think that you can play the violin, and if you do, that you have bought a new instrument made by a good maker. If you had, and found that over a few months or a year or two of playing the sound did not change, then your experience is at odds with the vast majority of violinists.

Are you saying that the Strad project showed that similar changes would be brought about by simply storing the new instrument? Or are you saying that it showed that the sound of a new violin does not change over time, with playing? If the second, sorry, but that's rubbish. The first... I'm very doubtful but somewhat more open to the possibility. However the assumptions of the study sound a little dubious to me because no two fiddles sound exactly the same, anyway, even ones made by the same maker from the same wood.

Of course everybody is grateful for many of the products of science and engineering. That doesn't mean we all believe that engineers and scientists can solve every problem, even if they are so sure they can. And excuse me, but many harmful products of technology are not being foisted on us by politicians, but are being developed by scientists who are either out to make money, or just want to see if something can be done, and to hell with the precautionary principle and the consequences in both cases. Unfortunately most of the rest of humanity hasn't the sense to object to many of these developments, but adopt them blindly, and those that do object are pretty powerless to do anything about them.
Title: Re: Soundboards again
Post by: oggiesnr on October 05, 2011, 07:35:49 AM
The OP asked whether the theory that the sound board material affects the sound had been properly tested.  It seems that it hasn't been although there is a wealth of anecdotal evidence that it does.

The key word however is "properly".  To do this requires a controlled experiment with repeatable parameters and measurable results.  In other words a scientific approach to take out ALL the variables other than the one being tested.  That is the only way to get a proper answer to question.  Other approaches may appear to work (build two boxes etc) but are not scientifically valid as there could be other factors at work (has anyone else had the experience of someone else playing your box and it sounding a hell of a lot better than when you played it?)

The "ageing" of stringed instruments is an interesting case in point.  Anecdotal evidence suggests it happens but the process by which it happens is not clear, is it the playing that is important or simply the materials bedding down with age.  I suspect that the strad experiment referred to above has not yet been going on long enough to be definitive.

Steve

Title: Re: Soundboards again
Post by: Anahata on October 05, 2011, 08:42:45 AM
Ian, I don't think that you can play the violin, and if you do, that you have bought a new instrument made by a good maker. If you had, and found that over a few months or a year or two of playing the sound did not change, then your experience is at odds with the vast majority of violinists.

There is an obvious alternative explanation. This was suggested to me many years ago by a friend who happens to be a very good violinist as a well as a physics PhD - and that is after playing the same instrument for months the player learns how to get the best sound from that instrument. He was actually talking at the time about the difference between a good violin and a good folk fiddle - a "fiddle" produces an acceptable sound quite easily but doesn't have very much expressive range - a top class violin is not always easy to play at first and the player needs to spend some time getting to know it before its full musical capabilities are revealed, which is saying more or less the same as you have stated above, but with no suggestion that it's the instrument that has changed.

For a control experiment, you need a first player who tries the instrument when it's new, then doesn't touch it while the second player works hard with it, then the first player has to come back and try the instrument again and decide if it's changed. Not reliable, though as it relies on long memory of something unquantifiable.
Title: Re: Soundboards again
Post by: Chris Ryall on October 05, 2011, 09:12:17 AM
Not sure about 'obvious'  but I think you're right. To get emotion from an instrument you actually need some instability. The most stable instrument possible is a sine wave generator.  Music it ain't. 

The really nice boxes I've played 'do what you ask them to do'  :|glug
Title: Re: Soundboards again
Post by: Stiamh on October 05, 2011, 09:38:43 AM
Good violin makers will tell you that they can easily make a violin that sounds very impressive and responsive when new, by making the plates thinner. They don't do this because they know that in time the sound of such an instrument will "collapse" or "wimp out" (I've heard different makers use both terms). They deliberately build instruments in a such a way that the sound is less impressive at first but will develop over time.

Anahata, it's quite true that a good violin will teach you how it needs to be played. I don't think that is the whole story though.
Title: Re: Soundboards again
Post by: Andy Simpson on October 05, 2011, 10:37:19 AM
An experiment that I'd find more interesting would be to build three identical instruments and have one played extensively for some time whilst leaving the other two under proper conditions, both structurally complete and finished but one of them with strings fitted under full tension and the other not strung up. Then do a proper blind test of all three and see if there's a difference.

From my own experiences as somebody who has "given birth" to some stringed instruments and seen the sound change noticeably in the first few days after being strung up with little playing, I'm inclined to believe that it's more to do with the instrument being under tension over time than being played.
Title: Re: Soundboards again
Post by: oggiesnr on October 05, 2011, 10:55:19 AM

The really nice boxes I've played 'do what you ask them to do'  :|glug


... opposed to what you thought you asked them to do  ;D

I used to have an old Gibson mandolin, beautiful instrument, the only problem was it was totally unforgiving, every little error (catching another string with the pick etc) would sound out in all it's glory!

Steve
Title: Re: Soundboards again
Post by: blafleur on October 05, 2011, 11:22:40 AM

The really nice boxes I've played 'do what you ask them to do'  :|glug

My accordion does what it's asked, it's my fingers that don't.
Title: Re: Soundboards again
Post by: IanD on October 05, 2011, 12:00:55 PM
An experiment that I'd find more interesting would be to build three identical instruments and have one played extensively for some time whilst leaving the other two under proper conditions, both structurally complete and finished but one of them with strings fitted under full tension and the other not strung up. Then do a proper blind test of all three and see if there's a difference.

From my own experiences as somebody who has "given birth" to some stringed instruments and seen the sound change noticeably in the first few days after being strung up with little playing, I'm inclined to believe that it's more to do with the instrument being under tension over time than being played.

This was precisely the point that the "twin sisters" experiment appears to show, which is why only 2 were made -- not claiming  that there isn't any change over time, but that the change is nothing to do with the amount that the instrument is played but is just due to aging, either due to material changes after manufacture (e.g. changes in wood moisture content) or stress relief.

So in answer to Steve Jones's objections, I'm not saying that there is no change over time, I was talking about the widely-held belief that this is due to "playing-in". On the objection that no two instruments are exactly the same, as far as I remember all Steve's objections were accounted for -- this is why they were "twin sisters" not triplets or anything higher; all the pieces of wood were mirror-image pairs cut from the same planks so the grain was exactly the same, the instruments were made as physically identical as possible, measured to have very little difference in their frequency responses, and then in a carefully controlled test it was shown that no difference could be detected -- so I'd say that according to the human ear, they were identical.

When comparing after some time the setup of each instrument was also redone by the same luthier before listening to make sure that any differences due to this were eliminated. I think that the conditions seen by the "live" instrument (humidity etc) were even tracked and the environment of the "reference" one was adjusted to track this. The listeners included luthiers, musicians and music critics to make sure that nobody could object that the listeners were not discriminating enough.

In response to the "they weren't played for long enough" argument (also anticipated), the experiment went on for several years (and may still be going on, I don't know) and most of the proponents of the "playing-in" theory are happy to agree that the big changes to new instruments happen in the first few months and the sound settles out after a few years at most.

In other words, the comparison was done carefully by people who knew what they were doing (I think it was associated with Texas A&M University who have contributed a vast amount over the years to the knowledge about musical instrument acoustics, specifically violins) and the results seem conclusive -- "playing-in" does not affect the sound of a violin.

I'll see if I can find the paper describing the details because it's really quite interesting, but it's years since I saw it and it may have been somewhere like the Science Reference Library and not findable on the internet.
Title: Re: Soundboards again
Post by: jb on October 05, 2011, 12:12:44 PM
here's a relatively recent study (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.169.5395&rep=rep1&type=pdf)
Title: Re: Soundboards again
Post by: IanD on October 05, 2011, 12:18:39 PM
here's a relatively recent study (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.169.5395&rep=rep1&type=pdf)

Excellent -- that's the paper I meant! I probably couldn't find it because I was searching through Navgary's material and Texas A&M.

It does show the care that has to be taken to be sure that there is (or isn't) a small difference due to some factor :-)

Ian
Title: Re: Soundboards again
Post by: IanD on October 05, 2011, 12:47:48 PM
Ian, I don't think that you can play the violin, and if you do, that you have bought a new instrument made by a good maker. If you had, and found that over a few months or a year or two of playing the sound did not change, then your experience is at odds with the vast majority of violinists.

Are you saying that the Strad project showed that similar changes would be brought about by simply storing the new instrument? Or are you saying that it showed that the sound of a new violin does not change over time, with playing? If the second, sorry, but that's rubbish. The first... I'm very doubtful but somewhat more open to the possibility. However the assumptions of the study sound a little dubious to me because no two fiddles sound exactly the same, anyway, even ones made by the same maker from the same wood.

Of course everybody is grateful for many of the products of science and engineering. That doesn't mean we all believe that engineers and scientists can solve every problem, even if they are so sure they can. And excuse me, but many harmful products of technology are not being foisted on us by politicians, but are being developed by scientists who are either out to make money, or just want to see if something can be done, and to hell with the precautionary principle and the consequences in both cases. Unfortunately most of the rest of humanity hasn't the sense to object to many of these developments, but adopt them blindly, and those that do object are pretty powerless to do anything about them.

Speaking for myself, the company I work for refuses to touch anything ethically dodgy with a very long bargepole, and I personally do the same. The problem is that you can't put the genie back in the bottle, a discovery made once is likely to repeated elsewhere even if swept under the carpet. And it's often impossible to know in advance to what purpose any discovery will be put -- are you suggesting that all scientific research should be closed down to prevent the nasty consequences, ignoring the good ones?

I suppose you'll quote the atomic bomb as an example of science gone mad. Well, in the past 66 years nuclear science killed about 200,000 people at Hiroshima and Nagasaki, and all the peacetime nuclear accidents including Chernobyl are estimated to have killed less than five thousand people -- all of which are tragedies. On the other hand replacing coal power stations with nuclear ones has saved the lives of tens of thousands of miners, and peaceful uses of radiation (CT scanners, radiotherapy, smoke detectors) have saved the lives of millions. However few of the positive factors make headlines in the same ways as the negative ones -- but you can't concentrate only on the negative consequences of something without also looking at the positive ones.

A perfect example is the Fukushima disaster, the latest estimates of which are that precisely zero people will die due to radiation effects, in spite of the worldwide press furore leading to calls to ban nuclear power -- which will increase the worldwide death toll since all other available energy sources which can replace it in the short term are more dangerous. Meanwhile hydroelectric power is seen as a clean green safe alternative, but dam collapses have killed tens of thousands -- in fact way more than the peacetime nuclear death toll.

How many of the world's ills are caused by unbridled and unregulated pursuit of short term riches -- and to hell with the long-term consequences -- is another problem entirely, but I suspect that scientists are still the wrong people to take the blame for this since they very rarely make all the money...
Title: Re: Soundboards again
Post by: Stiamh on October 05, 2011, 02:11:48 PM
The study is very interesting, thank you Ian and jb.
Title: Re: Soundboards again
Post by: Stiamh on October 05, 2011, 02:30:23 PM
Are you suggesting that all scientific research should be closed down to prevent the nasty consequences, ignoring the good ones?

No, what makes you think that? 

Quote
I suppose you'll quote the atomic bomb as an example of science gone mad. Well, in the past 66 years nuclear science killed about 200,000 people at Hiroshima and Nagasaki, and all the peacetime nuclear accidents including Chernobyl are estimated to have killed less than five thousand people -- all of which are tragedies. On the other hand replacing coal power stations with nuclear ones has saved the lives of tens of thousands of miners, and peaceful uses of radiation (CT scanners, radiotherapy, smoke detectors) have saved the lives of millions. However few of the positive factors make headlines in the same ways as the negative ones -- but you can't concentrate only on the negative consequences of something without also looking at the positive ones.

200,000 people dead from nuclear weapons. So far, Ian, so far. The potential for millions to be exterminated by them is still with us.

Actually nuclear power provides a pretty good example of scientists foolishly assuming that they can solve any problem. When power stations started to produce nuclear waste, we were reassured that science would soon find an answer to the problem of neutralizing the stuff, or at least disposing of safely and cheaply. Last time I looked, and this is what, 50 years on, no such answer has been found, and increasing amounts of nuclear waste are being shuffled round the world on trains travelling through heavily populated areas, and on ships.

The thing that nobody seems to talk about with regard to nuclear power stations - the elephant in the room - is that by building them you are condemning future generations to deal with them. Which assumes that the technical, financial and human resources to do so will always be there. A very big, and stupid, assumption, in my view.

Anyway... I think we are getting a long way away from soundboards in accordions.


Title: Re: Soundboards again
Post by: Chris Ryall on October 05, 2011, 03:03:51 PM
On the other hand replacing coal power stations with nuclear ones has saved the lives of tens of thousands of miners, and peaceful uses of radiation (CT scanners, radiotherapy, smoke detectors) have saved the lives of millions.

It's never that simple. We radiologists do a lot more good than harm with our linear accelerators and scanners - odd ratios are well, well toward 'benefit'. But a scan is still radiation and someone, somewhere gets a nasty as a result. eg the two scans I had in 2003 were equivalent to 'granite' dose from living in Cornwall for about a year. Interestingly, my motor bike would have been on Dartmore that very weekend 'had it not had its accident'  ::)

Radiation is funny stuff  >:E
Title: Re: Soundboards again
Post by: IanD on October 05, 2011, 03:46:33 PM
On the other hand replacing coal power stations with nuclear ones has saved the lives of tens of thousands of miners, and peaceful uses of radiation (CT scanners, radiotherapy, smoke detectors) have saved the lives of millions.

It's never that simple. We radiologists do a lot more good than harm with our linear accelerators and scanners - odd ratios are well, well toward 'benefit'. But a scan is still radiation and someone, somewhere gets a nasty as a result. eg the two scans I had in 2003 were equivalent to 'granite' dose from living in Cornwall for about a year. Interestingly, my motor bike would have been on Dartmore that very weekend 'had it not had its accident'  ::)

Radiation is funny stuff  >:E

I wasn't trying to say that nuclear technology was unequivocally a good thing (or that hydroelecticity is a bad one), simply that you have to look at all the pros and cons before making your mind up -- as David MacKay (author of "Sustainable Energy -- without the hot air") said, I'm in favour of numbers that add up :-)

What is clear is that the risks from radiation -- including Chernobyl -- are minute compared to the risks from many other everyday things which we accept without thinking. Falling while getting out of bed is the cause of death for about 100 people a year in the UK according to government figures, road accidents kill several thousand every year, but you don't hear many calls to ban beds or cars...

Ian
Title: Re: Soundboards again
Post by: IanD on October 05, 2011, 04:02:07 PM
Are you suggesting that all scientific research should be closed down to prevent the nasty consequences, ignoring the good ones?

No, what makes you think that?  

Quote
I suppose you'll quote the atomic bomb as an example of science gone mad. Well, in the past 66 years nuclear science killed about 200,000 people at Hiroshima and Nagasaki, and all the peacetime nuclear accidents including Chernobyl are estimated to have killed less than five thousand people -- all of which are tragedies. On the other hand replacing coal power stations with nuclear ones has saved the lives of tens of thousands of miners, and peaceful uses of radiation (CT scanners, radiotherapy, smoke detectors) have saved the lives of millions. However few of the positive factors make headlines in the same ways as the negative ones -- but you can't concentrate only on the negative consequences of something without also looking at the positive ones.

200,000 people dead from nuclear weapons. So far, Ian, so far. The potential for millions to be exterminated by them is still with us.

Actually nuclear power provides a pretty good example of scientists foolishly assuming that they can solve any problem. When power stations started to produce nuclear waste, we were reassured that science would soon find an answer to the problem of neutralizing the stuff, or at least disposing of safely and cheaply. Last time I looked, and this is what, 50 years on, no such answer has been found, and increasing amounts of nuclear waste are being shuffled round the world on trains travelling through heavily populated areas, and on ships.

The thing that nobody seems to talk about with regard to nuclear power stations - the elephant in the room - is that by building them you are condemning future generations to deal with them. Which assumes that the technical, financial and human resources to do so will always be there. A very big, and stupid, assumption, in my view.

Anyway... I think we are getting a long way away from soundboards in accordions.




Talk about cherry-picking your facts...

As far as I'm aware, not only has nuclear waste never killed anybody, but all the waste produced by all the reactors on the world could be crammed into one deep mine where it's unlikely to hurt anyone for a very long time. The problem is that nobody wants said mine to be near them. Yes it can't be guaranteed to be 100% safe for 10,000 years, but then what can be?

Before conjuring up the possibly deadly nuclear waste monster, you might want to consider what to replace it with. Not only all the certain -- not possible -- deaths that have resulted from fossil fuels, but the environmental impact of definitely dumping massive amounts of CO2 into the atmosphere -- or don't you believe that global warming exists either?

Or are Chinese miner's lives really worth so much less than those of anti-nuclear protesters, none of who have ever died due to nuclear waste?

You're right, this isn't the right forum to discuss this. However, as long as you keep coming up with irrational statements I just can't seem to resist the temptation to shoot them down with good old hard facts ;-)

Ian

P.S. Have you read the paper about violins yet? I know it doesn't agree with your views, but you could always still claim that you're right and a long carefully conducted scientific study by experts in the field is wrong...

P.P.S. Does this qualify as the new "thread drift" champion?
Title: Re: Soundboards again
Post by: Andy in Vermont on October 05, 2011, 04:03:07 PM
[...] as an engineer, I find it sad that people have views like this. [...]
Changing one variable at a time and seeing whether it makes a difference *is* science and engineering, and is exactly what I was suggesting -- but the results are only valid if you do the tests in a way that avoids observer bias.

As for the first statement, fair enough.  But you need not insult non-engineering approaches to improving quality of life.

Before the term "engineering" -- and the creation of the title "engineer" -- there were plenty of applications of the method that I described (which you claim for engineering) in the "techne" of farming and artisan-craft work.  

I understand the benefit of avoiding observer bias, but the problem with applying methods that attempt to eliminate observer bias is that the thing being made -- whether it is wine or a melodeon -- is ultimately observed, by an uncontrolled plethora of observers.
Add to that the notion that a given product may have a large degree of tolerance for inconsistencies in some characteristics and not in others. I think that a more approachable notion of techne is more appropriate to melodeon building, unless your goals are to replicate identical instruments, for instance, or to eliminate inefficiencies in production.  

I guess it boils down to whether you want to resemble Hohner (or some high-end version of Hohner) or Castagnari, and that is very much a matter of personal preference, and does not necessarily boil down to the superiority of engineering over non-engineering manners of acquiring knowledge and skill.  Apparently, what constitutes "properly" means something different for you and perhaps other engineers than it does for someone like Gaillard.  I'm sure that Gaillard is not attaining his results "randomly," but I also belief that he has not eliminated observer bias, or other biases, in a manner that would satisfy an engineer who works in a high-tech field.  I'm reminded of some woodworkers I know who make reproductions of Chippendale furniture etc who use incredibly accurate machinists' tools for measurement, and machines rather than "inefficient" hand labor in as many operations as possible. And yet, even if adjusted for inflation, their product is not as efficient, economically, as Chippendales, and Chippendale and his fellows never had access to machinery or measuring tools approaching that level of precision.

I believe that several of the accordion builders whom I admire have acquired knowledge about soundboard materials and their relative effects on sound (and other outcomes), and this is simply not available to the rest of us, except in the form of the decisions that the builders ultimately make, which may be as much, if not more, based on economics of the choice.
Title: Re: Soundboards again
Post by: IanD on October 05, 2011, 04:26:38 PM
[...] as an engineer, I find it sad that people have views like this. [...]
Changing one variable at a time and seeing whether it makes a difference *is* science and engineering, and is exactly what I was suggesting -- but the results are only valid if you do the tests in a way that avoids observer bias.

As for the first statement, fair enough.  But you need not insult non-engineering approaches to improving quality of life.

Before the term "engineering" -- and the creation of the title "engineer" -- there were plenty of applications of the method that I described (which you claim for engineering) in the "techne" of farming and artisan-craft work.  

I understand the benefit of avoiding observer bias, but the problem with applying methods that attempt to eliminate observer bias is that the thing being made -- whether it is wine or a melodeon -- is ultimately observed, by an uncontrolled plethora of observers.
Add to that the notion that a given product may have a large degree of tolerance for inconsistencies in some characteristics and not in others. I think that a more approachable notion of techne is more appropriate to melodeon building, unless your goals are to replicate identical instruments, for instance, or to eliminate inefficiencies in production.  

I guess it boils down to whether you want to resemble Hohner (or some high-end version of Hohner) or Castagnari, and that is very much a matter of personal preference, and does not necessarily boil down to the superiority of engineering over non-engineering manners of acquiring knowledge and skill.  Apparently, what constitutes "properly" means something different for you and perhaps other engineers than it does for someone like Gaillard.  I'm sure that Gaillard is not attaining his results "randomly," but I also belief that he has not eliminated observer bias, or other biases, in a manner that would satisfy an engineer who works in a high-tech field.  I'm reminded of some woodworkers I know who make reproductions of Chippendale furniture etc who use incredibly accurate machinists' tools for measurement, and machines rather than "inefficient" hand labor in as many operations as possible. And yet, even if adjusted for inflation, their product is not as efficient, economically, as Chippendales, and Chippendale and his fellows never had access to machinery or measuring tools approaching that level of precision.

I believe that several of the accordion builders whom I admire have acquired knowledge about soundboard materials and their relative effects on sound (and other outcomes), and this is simply not available to the rest of us, except in the form of the decisions that the builders ultimately make, which may be as much, if not more, based on economics of the choice.

I'm not insulting non-engineering methods to improve the quality of life, it's something that too little attention is given to. And just because craftmen didn't know the meaning of the term engineering (because it hadn't been invented) didn't prevent them improving things by trial and error, I expect there was a big heap of smashed-up failed violins behind Stradivarius's workshop.

You're still missing my point -- I've got absolutely nothing against craftmanship, or the emotional attachment to music and/or the instruments it's played on, and wouldn't want everything to be soullessly and mechanically churned out.

Maybe some accordion builders genuninely have knowledge that we can't share about many things including soundboards. Somehow I doubt that they've done the kind of experiments we've been talking about because that's not how most artisan instrument manufacturers work or think, but I could be wrong. Even so it's very easy for people to convince themselves of something when it isn't backed up by facts, even if they have years of experience.

All that I'm against is people stating *as fact* things like "the wood of a pallet-board has a strong effect on the sound quality" with no evidence to back this up.

If they precede the statement with "I think that..." and follow it up with "even though I can't prove it" than that's fine by me --it's no longer being presented as a fact, it's a belief -- like religion, but let's not get into that...

(people can hold what beliefs they want, so long as they don't claim that they're factually true)

The problem with the Internet is that claims like this stated *as a fact* get replicated and copied and spread around and can become accepted as received wisdom when they're nothing of the sort, they started out as one man's (or several men's) opinion.

The instrument will indeed be played to an uncontrolled audience, and everyone can agree that it is truly a fine instrument. But when somebody states that this is because of the resonance properties of the pallet board wood they'd better be able to back this up.

It's no different to cosmetic adverts saying we have a wonderful product which contains secret ingedient X and that this prevents wrinkles -- the Advertising Standards Authority would be down on them like a ton of bricks unless they can back up their claim with facts.

Ian
Title: Re: Soundboards again
Post by: Clive Williams on October 05, 2011, 04:31:39 PM
[[[ Good grief. That's quite enough; offending post removed; thread locked. Abusive posts will not be tolerated. -- Clive ]]]
SimplePortal 2.3.5 © 2008-2012, SimplePortal