Humidity balance in older guitars

kw: essays, musical instruments, maintenance, humidification

I have a few vintage guitars. One, a Takamine that I bought new in 1973 in California, became so dried out it began to come apart by 1980, so I unstrung it and stored it and bought another (which is now also vintage!), a Sigma Anniversary Edition. By then I was living in South Dakota, which is even dryer than southern California. I later learned of humidifiers for acoustic guitars, such as the one shown here. I got one for the Sigma.

The sponge can hold 20cc of water. In a closed and latched hard case it can keep a guitar from drying out for a month in pretty dry weather. However, it doesn't regulate the humidity, it just raises it, sometimes close to saturation, which can make a guitar body "fat", raising its pitch; it gets sharp and the tone changes, getting "hollow". And then, if you leave it outside the case to dry out, it goes flat. It's hard to find a balance.

Before my Mom died she sent me her guitar, the one I'd learned on some 60 years ago. She didn't know what brand it was, and the paper label was missing. Her dad had bought it used when she was very young. It had suffered some damage over the years, so I loosened its strings and just stored it for a decade or so. Then in 2016 I decided to take it to a luthier to see if I could afford to have it repaired and set up for playing.

The price was a bit steep, but affordable, so I had him go ahead with it. He did all kinds of things to it, and was also able to determine, from penciled and stenciled notations inside, that it is a 1905 Gibson Artist. He talked to me a long time about proper care once I took it home. He said a humidifier like the one I was using on my Sigma could ruin it, and recommended a 2-way pack such as the Humidipak by D'Addario.

The kit comes with three packs that contain a special mix of salts (probably mostly magnesium nitrate) and a gel, in an osmotic membrane. They can both raise and reduce humidity whenever it strays from 48%. The optimum humidity for an acoustic guitar is 50%, but anything in the range 45%-55% will keep it "healthy".

Two of the packs go in a 2-pocket bag and are hung between the strings; the other one is put in the neck compartment of the case. The box says each pack can release as much as 26cc of water in dry conditions. There is no indication of how much they can absorb when the environment is humid.

The luthier recommended that I get a room-size humidifier to use in the wintertime, to raise the humidity, and a digital hygrometer. From the readings on our thermostat, I already knew that the house humidity gets as low as 30% in winter. We run a humidifier in the basement, set at 50%. It regulates the whole house in summer, keeping it below 55%. Without it, humidity gets into the seventies and stays there for at least a couple months.

The humidifier I got needed daily attention, and even putting it in a closed closet with all the guitars, in their hard shell cases (by this time I also had a Fender 12-string), it needed filling daily. We were worried that all the humidity was going into the walls and ceiling and floor, and could induce mold, even though my digital hygrometer reported the closet's humidity stayed below 60%.

After a year of that, I thought things through and made a big change. The luthier had said that the speed of the Humidipaks to dry out humid air was slower than when they were humidifying dry air. I decided to do in-the-case regeneration of the Humidpaks, using the humidifiers I already had, and a few more I bought.

I made record sheets like this one, one for each guitar. I have a digital kitchen scale with a sensitivity of 1g (the weight of 1cc of water). I figured that if the Humidipaks can release 26g each, it would be OK for them to absorb 10-20g from the humidifiers.

Last October I began. I usually checked about twice monthly to see how things progressed. I would decide each time whether to add water to the humidifiers ("restoration packs"). The "+20" or plus-whatever records how much water I added. I used distilled water.

I noticed that each time I added water, the guitar would get a little fat after a day or two, then dry back out in another day or two. I figure that is the time it takes for the Humidipaks to absorb most of the water in the humidifiers. I didn't want to interfere too much by weighing daily.

Perusing this record, I can see that I had to add a lot of water from December through March. Middle and late February was warmer and wetter, so I tried not adding water for about a month. The humidifiers released their water quite slowly, and the Humidipaks kept their weight or even gained just a bit.

The bottom line shows measurements I made today. I added no water over the summer, after the final summer record on August 1. I only took this guitar out to practice, usually daily, because I used the 12-string to lead church singing. I had been using the Takamine for that the prior year or two.

I was told that Humidipaks will last about three years if a guitar spends most of its time in a closet with a small-room humidifier. Rather than burning through a gallon of water almost daily, I used about ten ounces of water for the whole year. This kept the Humidipaks a little over-full and the guitars all keep their tuning when not in use, and they only needed a little tweaking after being used for an hour or two. I'll find out if these last longer than three years after another couple of years.

Guitar Tuning - harmonics plus tweaking

kw: music, musical instruments, instruction

I played guitar for forty years before I obtained an electronic tuner. At first, I kept my guitar in tune by ear using the piano; my mother was a fine pianist and our piano was kept in tune. Before going off to college, I got a tuning fork, the A-440 one. I learned to tune the A string by sounding its 4th harmonic, striking and setting the fork, and tuning for zero beats. Then I tuned the rest of the strings using harmonics, which I'll describe shortly.

As my ear got keener, I noticed that tuning with pure harmonics led to certain chords, such as D, not sounding quite right, and I tinkered with the tuning a lot before I learned more about what I was doing. Finally I learned enough about equal temperament and how the harmonics relate to it to get my guitar tuned right, no tinkering needed.

This fingerboard diagram, the first octave cropped from a larger illustration in Wikipedia Commons, shows how the 3d and 4th harmonics relate everywhere but between the G and B strings:

The Nut is on the left, and the bridge would be off to the right a distance equal to the width of this image. The harmonics at the 7th fret (the rightmost set of orange dots) are the 3d harmonics, and those at the 5th fret are the 4th harmonics. You produce a harmonic by very lightly touching a string above the fret, being sure not to depress the string, and plucking the string near the bridge, preferably pretty close. The closer you pluck to the bridge, the greater the amplitude of higher harmonics. The 4th harmonic is strongest when you pluck 1/8th of the way up the string from the bridge, or about 3.2 inches (80 mm or so). This is also a good place to pluck for a strong 3d harmonic, which avoids having to adjust when you are trying to get two strings to sound on different harmonics so you can hear if they are in unison, or nearly so.

"Nearly so" is very meaningful in this context. This table shows why. Note first that the 4th harmonic of the A string is 440 Hz when it has been tuned to the tuning fork. Now, when the D string is in proper tune according to equal temperament, its 3d harmonic is almost half a Hz higher. By the same token, the 4th harmonic of the low E string is almost 0.4 Hz lower than the 3d harmonic of the A string, when the E is properly tuned.

Early on, I would tune to exact harmonics, using the A string to tune the E and D, then the D to tune G, the low E to tune the high E', then the high E' to tune the B. Working out the math, we find that with A at 110 Hz, the two E strings are then 82.5 and 330 Hz, the D is 146.67 Hz, G is 195.55, and B is 247.5 Hz. We'll see in a moment that tuning a pair of strings to harmonics this way results in the lower string being 2 cents sharp with respect to the higher string. A "cent" is 1/100th of the halftone interval. Thus if E and E' are 2c sharp, B becomes 4c sharp, while D is 2c flat and G is 4c flat. Thus the G-B interval is off by 8c.

One would think that these little errors, which amount to fractions of a Hz, would be too small to hear. Not so. An error of one musical cent or two is not easy for everyone to hear, but almost anyone can hear an 8c mistuning. This table shows the frequency and cent differences, string pair by pair, and how I compensate for them:

As you can see, for the lower strings, the frequency differences are less than 1Hz, so it is hard to adjust by counting beats. In my experience, I cannot hear the beat when it is less than 1Hz, and I can barely hear the B4 to E'3 beat. What I do, then, is adjust until there is a difference I can barely detect. For example, I first tune the low E to the A using E4 and A3, and get it exact by nulling the beats. Then, knowing that the E is a tad sharp, I flat it until I can just detect the difference. I do so for each pair. This procedure leads to a final tuning so close to equal temperament that the G and B sound good together, and the chords D, A, C and E all sound equally good.

Of course, now that I have an electronic tuner, I can get a guitar in tune pretty quickly, but I have found that the green light stays on over an interval of plus-or-minus a cent or two. Thus, after I tune electronically, all the strings are probably within a cent or so of being right, but if I tweak a little by ear, I can make my guitar sound just a little bit better.

One caveat. This does not take inharmonicities into account. Because of small nonlinear effects the harmonics are slightly sharp compared to the open note, but for a guitar using steel strings the difference is half a cent or less at the 4th harmonic, and negligible at the 3d harmonic. Nonetheless, the result is a slight stretch of the tuning. You may know that pianos are usually stretch tuned, and over the 88-key range, the highest notes are 10-15 cents sharp as compared with the lowest notes. Piano strings are stretched tighter than guitar strings, and the harmonics get sharper as a result. Stretch tuning makes notes a few octaves apart sound better together. This is also true for the guitar, in principle, but it is very rare to play notes more than 3 octaves apart on the guitar. I haven't found a noticeable difference.