Don Noon

300+ year old Spruce

Recommended Posts

I suspect I played the violin Melvin made from the same log you tested. He had the decency not to tell me there was anything funky about the wood until after I'd reached some conclusions of my own. 

 

I thought there was something intangible, and hence very very difficult to explain, in the core of the sound which was more familiar on older instruments. For a straightforward Guarneri model, it just seemed that you could work the strings a bit more to get more colours and it had the impression you could craft the sound in a way that was positively different (a guarded way of not saying 'better' but kind of meaning it), than I commonly expect on a new fiddle. I'm not sure that was a golden-bullet in itself as other variables could affect other elements of the tone in a way that could over-compensate that idea. But I suppose there was a sense of gravitas in the sound that was appealing, and more familiar within older instruments. It made sense to me when Melvin told me the age of the wood. However, these are utterly subjective. I didn't feel that there was an overwhelming argument for old wood, but I felt this violin - possibly because of the wood - contributed something pleasantly distinctive. I'd buy it! :) 

Share this post


Link to post
Share on other sites

If there were errorbands plotted around those two linear fits then I think that plot would look much less conclusive that old wood is significantly different than new. There look to be ~100 samples in each set and the fits show ~10% difference. Seems like the sets aren't big enough to draw any conclusions.

Share this post


Link to post
Share on other sites

Error bars - the 600 lb gorilla no one talks about.

 

Not just that. 

 

Correlation plots can be very sensitive to values at the ends of the distribution. 

 

I don't see anyone reporting sensitivity analyses; for example, where data points are removed progressively and then the residual 'correlations' are examined.

post-24474-0-23728500-1440339538_thumb.jpg

Share this post


Link to post
Share on other sites

However, about 20-30% of hemicellulose will degrade in 300 years, so I tend to agree with studies saying that elasticity in the radial direction will decrease a little. 

Seems like the sets aren't big enough to draw any conclusions.

Not just that. 

Correlation plots can be very sensitive to values at the ends of the distribution. 

 

It looks to me like there are enough points, regardless of throwing out a few at the ends of the distribution, to say that there is definitely a trend to higher stiffness/weight ratio (or speed of sound) in the old boards.

 

However, I'd be hesitant to say that these datapoints prove that the difference is due to aging, as there could be a systematic influence, such as old wood being cut from nice, old-growth forests, and modern boards being cut from fast-growing farmed trees.

 

The thesis concludes that the modulus is not correlated to age, but density is.  I'm not sure how that all got worked out, exactly, but I'd be extremely cautious about this, as it appears to me that the old boards might have just happened to be lower density than the modern boards, on average.  Nevertheless, the modulus/density ratio appears significantly different, which is the real issue, and perhaps paying attention to speed of sound instead of modulus would have been more enlightening.

 

I think it is important to remember all the things that can happen with age, and how it can skew results.  For example, hemicellulose degradation alone would decrease density and modulus, and likely reduce EMC which would further decrease density.  However, the wood also shrinks considerably, thereby INcreasing density and modulus. There can also be some polymerization of hemicellulose and lignin, firther increasing modulus.   It is not immediately obvious which effects will win out in the end for modulus, but I think that all of them tend toward a greater stiffness/weight ratio.

 

I look at hydrothermal processing as a possible indicator of what might go on with age.  In my results, moderate processing decreases density by ~7% (lower EMC is a major contributor to this), and modulus increases a few percentage points.  With stronger processing (increasing age?), density continues to decline, but modulus begins to degrade.  Damping declines constantly with the strength of the processing.

 

These percentage changes (except for damping) can easily be lost in the much larger natural variations in tree-to-tree properties, unless the sample numbers are large and well-controlled to exclude systematic differences.  In processing, before/after measurements can easily be made on the same board to factor out the unwanted variables.  It's not so easy trying to do that for 300 year old wood.

Share this post


Link to post
Share on other sites

It looks to me like there are enough points, regardless of throwing out a few at the ends of the distribution, to say that there is definitely a trend to higher stiffness/weight ratio (or speed of sound) in the old boards.

 

The problem lies in the "looks to me like there are enough points" - most of them are clustered around the middle. 

 

All I am saying is that it could be very instructive to take out 5% of the points at top and bottom for the old and the new wood, and then see how it looks and compare it to the correlation line draw by the computer.

 

Incidentally, I despise 'correlations' because they are so prone to providing a false sense of understanding.

Share this post


Link to post
Share on other sites

I think the only thing anyone needs to know is that softwood....Spruce, Redwood, Fir, Cedar are resin/sap laden....And that over a period of x amount of time these resins will loose moisture and dry out. When fresh, the resin is soft and wet, when dry is is rock hard and dry and that over time this will effect the way the instrument sounds.

Share this post


Link to post
Share on other sites

Old thread, but there is some more recent relevant test information that should be tagged on here.

EMC testing was done, primarily to see more precisely what difference hydrothermal processing causes, but the very old samples were included.

EMC.jpg.991ebe5a259d9b4128db7e73080b6411.jpg

Share this post


Link to post
Share on other sites

My experience is similar. Processed (heat treated ~3 h 160C) wood is more stable to humidity change than old wood.

It was a surprise to me because I thought that old wood would be much more stable 

My heat treated wood has max EMC ~7-8% in summer time ~60-70% RH, other wood including old is at the same time 10-11%

This is why I find Bruce's reaserch interesting.

Share this post


Link to post
Share on other sites
1 hour ago, Peter K-G said:

My experience is similar. Processed (heat treated ~3 h 160C) wood is more stable to humidity change than old wood.

It was a surprise to me because I thought that old wood would be much more stable 

My heat treated wood has max EMC ~7-8% in summer time ~60-70% RH, other wood including old is at the same time 10-11%

This is why I find Bruce's reaserch interesting.

Do you heat treat your wood in an oven or under IR lights or something else entirely?

Share this post


Link to post
Share on other sites
3 hours ago, Frank Nichols said:

Do you heat treat your wood in an oven or under IR lights or something else entirely?

In an oven starting from putting the wood in when it's cold and turn the oven on at 160 C. I take the pieces out every 30 or 40 min and weigh them and change their place in the oven when I put them in again. when I'm done I turn the oven off and leave it in the oven over night to cool down slowly. The two graphs are examples of baking in summer time and in winter

59dfc159280ea_2014-06-2117_47_00.thumb.jpg.ef457aeb8422e1e233758954f82d23b7.jpg

BakingGraph.thumb.JPG.29e2e83064b284f2027589a17d6bf440.JPG

Wood2017.thumb.JPG.3301810fc6e3a1f5105924809819a29d.JPG

Share this post


Link to post
Share on other sites
9 minutes ago, Marty Kasprzyk said:

Have you tried your oven baking on a violin?  

It would be interesting to see how much its sound changes.

That is scary, would the glue hold up? I assume one would remove the pegs, tailpiece, bridge and strings and only have the wood... (maybe leave the fingerboard on?)

 

Share this post


Link to post
Share on other sites
20 minutes ago, Frank Nichols said:

That is scary, would the glue hold up? I assume one would remove the pegs, tailpiece, bridge and strings and only have the wood... (maybe leave the fingerboard on?)

 

Or just do the wood before you make the violin .....

Share this post


Link to post
Share on other sites
16 minutes ago, Marty Kasprzyk said:

OK, I agree it might not be a good idea to try it on a violin.  

 

How about trying it on a viola?

:)

20 minutes ago, puckfandan said:

Or just do the wood before you make the violin .....

Well, the idea was to see what change occurred in the sound, so, it kind of would need to be playable before and after baking :)

Share this post


Link to post
Share on other sites
On 11/24/2014 at 5:18 PM, Don Noon said:

[...]

Come to think about it, the wood in trees can be extremely old (especially in the middle) even before they are cut down.

[...]

!!!!   Never thought about it that way before.  But so true!

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.