Official Luthiers Forum!

Most posts I see here and advertisements from most luthier sites tout Adi, Lutz, European, and lastly Sitka spruce as choices for soundboards. It’s rare to hear or see Engelmann spruce being used except by mass producers. Is Engelmann an inferior choice or just not trendy?

I've used Engelmann. It can be a bit soft, and some experience it as having less headroom. But the guitar I made with it came out very nice.

I am getting to the point where I worry a lot less about the specific type of spruce and focus a lot more on the characteristics of that particular top set.

I use (and have) a lot of Engelmann. It is a very variable species. It can be paper white or have color and be very "striped". It can be as dense and stiff as red spruce, or lighter than Norway (German) spruce. It tends to come from smaller trees that often grow with spiral twist. This means you can have "no runout" on one edge and major runout on the other. Or the runout may be uniform across the piece (no twist) and from none to terrible, as in other species. Engelmann hybridizes with a number of other spruces, which may account for some of the variation.
With Engelmann I think you have to be selective about which piece you use for which type of instrument, and also which edge of the top you join. I select the edge with the least runout, rather than the width of the grain lines. Generally less runout = stiffer, and has other advantages.
I have and use "A" grade tops, and "3A" tops and although the "3A" tops are cosmetically better., the average runout was about the same as the lower grade tops. I've noticed that as the wood ages the color differences of the lower grade tops becomes less visible.
My suggestion would be to either hand select the Engelmann you buy for the qualities you desire (if you build a particular type of instrument) - Or buy a quantity of random soundboards and build a number of different types of instruments they will be a good match for (which is what I do).

Engelmann is so 90's

I've used it and probably have at least ten Engelmann tops in my collection. It, at least used to be, was always very uniform in color and grain so it made for a nice looking top. In the end though it's just another spruce.

I've used it on some archtops and liked it. I really tend to prefer somewhat colorful Sitka tops aesthetically, but that's just me. Nothing at all wrong with Engelmann in my book.

Dave

IMO spruce is spruce. If you measure the properties of a lot of tops, you'll find that the long-grain stiffness at a given thickness pretty well tracks the density, with all
softwoods following the same rule. The cross grain stiffness correlates most closely with how well quartered the top is. Cedar and redwood generally have far lower damping than the spruces, so they're different in that respect. The average density does vary by species, but there's so much variation in each species that there's plenty of overlap. Some people maintain that Sitka is the 'odd man out', differing from the other spruces, but so far I have not been able to find any consistently measurable difference. Even 'identical' guitar made with 'the same' wood sound different, so I think it's risky to attribute too much to the species.

Engelmann does seem to average somewhat lower in density than the other spruces, but, again, it varies a lot. Some of the densest tops I've gotten have been labeled 'Engelmann'.

High density in the top wood tends to end up producing a heavier top, and that usually seems to favor 'headroom'. Lower density, lighter tops tend more toward 'responsiveness'.

Factories tend to work to the average, since they can't take the time to actually measure each top. Factory guitars thus tend to vary a lot, but reflect the average properties of the wood for the most part. "The race is not always to the swift, nor the battle to the strong, but that's the way the smart money goes".

An individual maker can work 'to the wood', choosing a top that is likely to produce the desired sound and bringing out the best in it, no matter what the species.

My next two Parlors will have Engleman tops. Will be my first try with that species, but I'm expecting it to be an advantage for these smaller bodies that really need as much warmth and responsiveness as I can get. My jumbos use Adi, and are massive sounding. If I get a player who wants the ultimate in headroom, I leave the Adi top a little stiff, and they are so loud you can't hold a conversation when playing it at full volume.

I don't have much experience, and nothing yet on guitars except for a baroque that can't keep its bride on, and a curly redwood arc top; but I agree with everyone. Wood is not uniform. I bought an example of wood from Simon Chambers when I started making violins. I had made a couple, and wanted real wood. It was 3 pieces of Englemann, from .3 to .45sg. I don't know what the others did; maybe I messed them up, I don't remember, but the ,3 one I made into a Montagnana violin with a very high arch, and a very stiff curly birch back. I don't play, but it plays very easily.

With guitars, shaping the sound would be influenced by the bracing, so a less stiff piece could be quite a bit thicker, or the bracing more substantial, or some combination; it is up to you. On my violin, I think that the stiff back pairs with the high arch to make the whole thing work.

Alan, I've. heard lots of talk about damping before, and I've never heard an explanation of it, or I didn't understand what they were saying. Is it something like a mute, or a blanket, that has a tendency to stop the wood from vibrating, or ringing. My redwood arch top with nylon strings will ring like crazy on some notes, and chords, and on others it is dark sounding. I like that. I noticed that the ringing is mostly from the low E and A strings. Would that be an example of low damping? The bridge just wants to keep moving?

_________________
Why be normal?

Damping is the rate at which vibration energy is dissipated within a material or structure. You could think of it as some sort of 'friction' that saps energy from the system and turns it, ultimately, into heat.

There are lots of different causes of damping, and they can behave in different ways. Usually damping will dissipate a certain proportion of the energy in the system per cycle of vibration, independent of the frequency. Think of it as interest on a loan with a floating balance: the more money you borrow the the higher the interest payment. Damping is often measured in terms of a 'quality factor' or 'Q value', which is the proportion dissipated per cycle. A piece of spruce might have a Q value for bending along the grain of 120, and for cross grain bending maybe 65. That is, if you took narrow strips of the wood with the grain running along or across, and started them vibrating with a certain amplitude, the long grain strip would dissipate 1/120th of it's energy (about .83%) per cycle, and the cross grain strip 1/65th (about 1.5%) per cycle. Cross grain bending in this case 'wastes' a lot more energy than long-grain bending. In a guitar, of course, you always see the two together.

In my measurements, I often see redwood and WRC with long grain Q values closer to 200.

For most materials we can talk about 'the' damping factor: it's pretty much the same at any frequency. This may not be the case with wood. One person who did alot of measurement (Haines) found that for most softwoods the damping factor was low (and the Q high) at low frequencies, but started to increase rapidly above 2000 Hz. In some Sitka samples he found that the low frequency damping was higher at low frequencies, and dropped off slowly until you hit 2000 Hz, after which it rose in the same way as other spruces. He could think of no reason for this, and was not at all certain that any of it was 'real'. It's possible that the frequency dependence was an artifact of the way he measured his samples, but that still begs the question of why Sitka would be different. It's a lot of work to check this stuff out, and so far as I know, nobody has put in the effort since.

This all gets complicated, as do most things associated with guitar acoustics. We're not entirely sure what effect damping gas on the perceived sound of the guitar, for example. It's almost impossible to isolate it as a factor in real guitars, given that even 'identical' guitars made from 'the same' wood will sound different. The damping in a completed guitar is a function of the whole structure, and there are lots of things other than the material properties that would contribute. It would probably be hard to 'prove', but I suspect that it would be hard to get the structural damping below that of the material, but hard to avoid having it higher. In other words, it's easy to mess up a good piece of wood. ;(

I hope this helps.

Alan,

Can you outline a process for measuring Q?

I’ve had wonderful results with Englemann — the first on an OM surprised the heck out of me. It is softer than other spruces — will dent if you look too hard and dark sanding dust embeds itself easily.

There are two ways I know of, but I'm only really familiar with one. The one I don't use is based on counting the number of cycles it takes for the vibration to drop off by a certain proportion. You'd record the sound in some way and look at the wave form.

I use a 'Q by band width' measurement. I already have the setup for it, and since I use a vibration test to determine the other properties this is easy to do, although it takes a little time. The idea behind this is that you only have energy being dissipated when the thing is moving, and 'loss' is proportional to how much it's moving. Any system is, by definition, easiest to drive at a resonant frequency. If the 'losses' are low, then you should be able to drive it to a higher amplitude at resonance, but anything is harder to drive off resonance. A high Q system, then, will show a sharper resonant peak, and you can measure that 'sharpness'.

We're all aware of this relationship, even if we don't think much about it. A material with low losses, like aluminum, rings for a long time when it's tapped, as you'd expect. It also tends to have a well defined pitch; all of the vibration is happening within a very narrow band of frequencies. Something like styrofoam or cardboard dissipates the energy faster; it has a poorly defied pitch, and dies out quickly.

What you need for this test is a way to see how the amplitude of vibration around any resonant frequency drops off as the frequency is changed. If it drops off fast on either side of the peak the Q value is high. There are two ways you can do this.

One is to use a spectrum analysis program, such as the one included in 'Luthier Lab'; a free Android app that was written by a friend of mine. Any 'FFT' program will do. You find the best spots to hold and tap a piece of the material; the ones that give the clearest, strongest, and longest lasting sound when you tap it. For something like a half a guitar top you'd hold it about 20% of the way in from the end, along one edge, and tap in the middle or at the very end. You record the sound, and use the 'frequency analysis' module to produce a spectrum. There should be one tall peak at a low pitch that is well defined; more or less like the Gaussian curve you used to ask your teacher to use in grading. Find the frequency peak, and the frequencies on either side of it where the energy is one half what it is at the maximum. If the program displays the amplitude as decibels, then you want the points on either side of the peak that are three dB lower. If the amplitude is on a linear scale, such as, say, the output of a microphone or pickup, then you find the peak value and go down to where the amplitude is .707 (1/2 the square root of two) of the maximum. The energy in the system goes as the square of the amplitude, and you're looking for the points where it's 1/2 the maximum.

Now you've got a peak frequency (F-p), and the points on either side of it where the energy in the system is half of what it was at that peak (F-hi and F-lo). The equation for Q is:
Q= F-p/(F-hi -F-lo),
that is, it's the peak frequency divided by the band width.

I do it a little differently. I use a vibration test to obtain the Young's modulus (E) of the wood as well a the Q value. The E value is calculated from the peak frequency, along with the mass and size of the piece, so it's easy enough while I'm at it to find the Q. The tophalf is supported on foam blocks at the stationary 'node' points, again, about 1/5 of the way in from either end (or side, for cross grain measure). I use a loudspeaker and signal generator to drive the plate with a tuneable sine wave signal (which can also be provided by Luthier Lab), and some glitter to show the pattern of node lines when I've hit the peak. I use a small piece of iron stuck to the end of the plate and a magnetic pickup, to read the amplitude, using a DVM. If you have one, a meter that can give mA-AC output readings with a needle would be easier to use, since it's easy to pick off the half power points directly, instead of having to calculate them from the peak output. Note that once you've found the frequency peak you don't touch the power knob on the signal generator. The idea is to find the points on either side of the peak where the energy is 1/2 the max with the same driving power. Again, once you have the peak and the two down points on either side you plug the values into the equation given.

There are pitfalls in this, as it any measurement. Either way you do it, you have the Q value at a particular frequency. We assume it's the same, or, at least, similar relative to other pieces of wood, at other frequencies, but there's no guarantee of that. Haines used to use long strips of wood, and run several different (I think three) vibration modes, to get Q values at different frequencies, He'd then cut the strips shorter, and do it again to get a family of higher frequencies. That's OK if you're not planning to make something of the strips afterward.

We're trying to measure the E and Q values of the wood for bending along and across the grain separately. When the wood is bending purely along or across the grain the stationary node lines will be straight, and parallel to the ends or edges. If those lines are curved then it's bending in more than one way at a time, and numbers you are getting reflect that mix of bending. I had one Sitka top a while ago that had nice straight node lines that ran on a diagonal across the top, and the lengthwise E and Q values were both lower than expected, particularly the E. I re-cut the piece a little smaller, but still big enough to use, and it tested out much better, ending up being a good top. This is one of the advantages of using the method I do; I can see if there might be a problem with the modes, and maybe even do something about that. A tap spectrum should show a distorted peak if there's another mode that's close enough to affect the one you're looking at, but it may not.

OTOH, if I put my hand too close to the plate as it's vibrating it alters the pitch, and especially the loss factor. I try to get the plate up off the bench far enough to minimize this, but....

One person I know was using a spectrum analysis program on his computer for this, and forgot to turn off the computer speaker. Feedback was driving the plates a bit, and he was getting some phenomenal Q values. And so it goes.

Here is a side by side comparison of two Engelmann soundboards. They are quite different in grain and coloration. The fine grained "white" one is less dense and lighter in weight than the one on the right that has wider grain and more color. They both have little runout and are well quartered, and should make decent sounding instruments if used properly. The white one is what people seem to expect and would be graded higher, but for larger instruments I have more luck with the denser, stripy top wood.

Curious about compression grain wood. About 25 years ago I bought some wood from Don Musser. Among that were two sets of compression Blue(?) Spruce. Compression came because the tree had tipped diagonally in its early life, and the compression was the tree's effort to strengthen itself. Don thought this was prize wood, but I have seen some posts negative to compression grain.

Opinions please.

I found this in response to a similar question on AGF in 2015. It is a verbatim response submitted by Alan Carruth.

“ There's 'compression wood, and 'compression wood'...
Some of the suppliers talk about 'compression GRAIN'. This seems to be wood that comes from low down in a larger tree, and has heavier than normal late wood lines. You see it quite often in redwood and Doug Fir, for example.

'Compression WOOD' forms in softwoods when there is an asymmetric load on the tree. This can be due to the tree leaning, or losing a branch, and is often seen on the south side of trees (in the northern hemisphere) because they have more branches on that side. It forms, as the name implies, on the compression side of the tree.

Both of these are a response of the tree to the need to support a heavy load. In the case of compression 'grain' the load may be more or less uniform, so the hard ring lines might be seen more or less all the way around (correct me if I'm wrong, John Arnold). It's simply the weight of the tree, plus any built-in stress that it needs to take into account. Wood like this tends, in my experience, to have relatively high density in relation to it's stiffness, but the damping (the 'ring' when it's tapped') is not impaired.

Wood like this might well be desirable for some uses. When worked to the correct stiffness it makes a top that is heavier than usual, which seems to help out with 'headroom'. I'd look for wood like this in particular if I were making a Dread with scalloped bracing.

Compression 'wood' is a specialized beast. Wood 'cold creeps': if you put a bending load on a piece of wood over along time it deforms and takes a set in that direction, even when the load is very far below the nominal yield point of the wood. That's why we use truss rods in necks, for example. Trees have not evolved built-in steel rods, and have trouble resisting a bending load that is continued over the long term. To fight it they produce 'reaction wood'; wood where the fibers run on diagonals in an effort to convert shear loads into tension and compression within the fibers. Hardwood trees make 'tension wood' on the top of a branch to try to hold it up, while softwoods make 'compression wood' on the under surface to do the same. Reaction wood has a lot of built in stress, since that's what's holding the thing up. It tends to move a lot when you cut it, and more when it's being seasoned. Because of the interlocked cell structure it's hard to get a smooth cut, and doesn't finish well. It's basically the reason we don't usually use wood from branches to make things. As I said above, there are a number of reasons why you might get compression grain in the trunk of a softwood tree, and sometimes (often) it will 'come and go' as a response to changes in the life history of the tree. If a big storm takes off a large lower branch you could get compression wood on the other side of the tree until the symmetry has been restored by new growth, for example. Trees react very quickly to these sorts of stress, according to one reference I have. Thus you could have a 'racing stripe' of compression wood for a few years that fades out.

The interlocked grain of compression wood in general has higher density than 'normal' wood. Because of the fact that the fibers don't all run along the axis of the tree the stiffness along the grain is lower. It also tends to have much higher damping than normal wood. These characteristics, along with the lower stability, and difficulty of working and finishing, mean that we tend to avid it when possible. You can make a top with it, but why bother?”

Alan did ask me to correct him, so here I go. I stick with the scientific literature here, and augment it with my experience of cutting dozens of red spruce trees and the resultant multi-thousands of guitar, mandolin, and violin tops.
Alan insists on relating compression wood to the lower part of the tree, but that is simply not the case. Compression wood (heavy latewood lines) occurs anywhere the tree is subjected to unusual compression loading. I say unusual, because incidence of compression wood is not more prevalent at the butt. When the compression is due to leaning, it is always on the side of the log toward the lean. It also occurs underneath large limbs. In my experience, compression never occurs all the way around the circumference, and that bolsters the idea that the normal weight of a large tree does not trigger the formation of compression wood at the butt. If you are looking for minimal compression wood, find a tree that is straight, on level ground, that is not subject to high winds or deadfall around it.
Conflating compression wood with interlocked grain further confuses the issue, since they are competely independent. True compression wood rarely contains interlocked grain. Interlocked grain IS usually confined to the lower part of the tree, if it is present at all. Nowhere in the literature is interlocked grain (or curly stumpwood, for that matter) called 'compression wood'.
I have used both compression wood and interlocked grain, with great results.
All through the 1980's, I used Engelmann almost exclusively, and the best results were always from using compression wood. I sourced most of my Engelmann from construction lumber, so all the compression tops had (thankfully) not been weeded out by a tonewood cutter.
Steve Gilchrist has bought red spruce mandolin tops from me on numerous occasions, and he always buys compression tops.
Why bother? Because it can make great instruments, as Clay S. has indicated.

_________________
John

Another supplier I know has said that he sees more compression grain further down in the tree, and I was basing my posts on that.

I'm stymied by the restriction of the use of 'interlocked grain' to 'stripe' figure. Reaction wood that I've run into has had a 'cross grained' structure that is certainly 'interlocked' on a fine scale. So...?

I'm just trying to make sense of what I've seen, and read, and what's been reported to me.

Heavy late wood lines, as in Clay's right hand sample, tend to add mass faster than they add stiffness along the grain. If you work those two tops to the same long-grain stiffness the denser one will end up a bit heavier. This gives it different acoustic characteristics, in particular seeming to help confer 'headroom'. If that's what you're after in your guitars or other instruments, that's the wood to use.

When I split compression wood, it almost always splits flat, no ripples across the grain. That tells me that there is no interlocked grain. The property attributed to compression wood (lower long grain stiffness per unit density) can be tied to the higher lignin content (which contributes more mass than long grain stiffness), and a correspondingly lower cellulose content.
Wikipedia:
"Compression wood has a higher proportion of lignin than normal wood. Compression wood has only about 30% cellulose compared to 42% in normal softwood. Its lignin content can be as high as 40%."

While the long grain stiffness is less, I consistently see a higher cross-grain stiffness when the tops are worked to 0.125". That is where I concentrate my design parameters, and it seems to work out quite well.
Gilchrist prefers compression tops because he contends they are more consistent. I preferred compression Engelmann because it is harder and stiffer across the grain......less like WRC and more like the other spruces. Most of the compression Engelmann I used is pretty indistinguishable from red spruce.

_________________
John

Delete

So how do you identify compression wood then?

+1

Interesting stuff.

Dave

I never made a conscious effort to buy Engelmann tops with compression wood. It comes as run of the mill when you buy lower grade tops. As John mentioned it is actually selected against in higher grade tops. Around 10 years ago, by buying in bulk (50-100 tops at a time) I could buy "A" grade dreadnought size tops for $3 a top +$2 shipping ($5 per top delivered). I expected they would have defects that would limit their use for dreds, but I build many smaller instruments and figured I could cut around many of the bad areas (knots , pitch pockets, knot shadow, etc.) and "upgrade" the wood in that fashion.
Although some of the tops did have those physical inclusions, most of the tops were downgraded for color, irregular grain spacing, and prominent late wood lines (compression). A few of the tops have terrible runout, but as I mentioned it occurred in about the same ratio as the "3A" tops I bought. I figured I could afford to lose a few and still be ahead. As I worked with Engelmann, I found I liked - some - of the "lower" grade tops better than the "better" grades for certain types of instruments.
Here is a list of characteristics this tonewood dealer goes by to grade his tops. Notice Compression increases as the grade decreases (among other things - Not all tops are downgraded for the same reason, so there is no guarantee the tops will have compression) Also notice there is no price listed for "A" grade, although the grade itself is noted. It is not a grade that is very marketable commercially, which is one reason it sold so cheap (at least back then). He was quite happy to give me a "bulk" price to clear out stock that was accumulating with little pecuniary value. You can probably still buy 12 to18 "A" grade tops for the price of 1 "Master grade" soundboard.
Price list/grading: http://www.kootenaytonewood.com/products

Thicker, darker latewood, which makes it resemble Southern yellow pine. Martin calls it 'racing stripes'. It often occurs as a few growth rings, or in bands that are no more than 1" or 2" wide, since it is a correcting mechanism for the tree. Once the tree straightens itself or reduces the stress, it will revert to producing normal wood.
It is this inconsistency in the appearance and properties that provide a challenge when using compression wood. Where the compression is a narrow band, I have selectively thinned that area to produce a more balanced flex across the grain.
Ideally, you want compression throughout the top, and that is what Gilchrist selected. For the reason cited, it is a bit easier to find on a 5 1/2" wide mandolin top than an 8 1/2" wide dreadnought top.

This is one of my last Engelmann guitars, with compression. 0000 size, 1990 build. This top came from a 2 X 10 from the local building supply.

_________________
John

Hi John,
That is a beautiful guitar. The shape reminds me of an old Stella 12 string I once had.
As the spring wood darkens it blends in better with the autumn wood and the racing stripe look lessens.

OK, I think I understand what compression wood looks like, but want some confirmation. Here's a guitar I'm working on with a Sitka top that's the kind of Sitka that I like, both visually and tonally. Is this compression grain? Thanks!

Dave