Accordion Revival



© 2014 George Bachich All rights recerved.

 Background information

        Why is there so little information available about repairing and tuning accordions? Why haven't experienced repairpersons and tuners shared what they know? Could it be because they want to guard the source of their livelihood? Could it be because they think repairing and tuning accordions is a trade that must be learned over a lifetime or through an old world style apprenticeship under expert supervision, and that sharing a little knowledge with novices would only get people into trouble? Could it be that even though they are good at repairing and tuning, they are not so good at communicating what they know? Could it be simply that there are so few of them and so few of us that they seldom meet anyone who wants to know what they know? If so, then they may sense that the market for their knowledge is so limited that it is simply not worth the time and expense to put what they know into a cohesive, coherent, communicable format.
        It seems to me that all of these reasons are valid, and that with so many good reasons for the existing information void, it could continue indefinitely if I don't do something about it. So I plan to fill that void, first with this website and later with a another book, which I am currently calling PIANO ACCORDION REPAIR MANUAL AND TUNING GUIDE.

        The information posted on this website will endure only as long as I continue living and paying the rent on the website server. When that ends, it will all disappear with a click of the server owner's mouse unless I put it into a more lasting format; hence the book idea. I don't expect the book to be a big seller volume-wise, because it will serve only this very tiny niche market that you and I occupy, but to make it even remotely possible that I will be able to recoup the expense of publishing it, I am reserving all copy rights to all the material posted here. Please use the information you find here to educate yourself and to rescue as many old accordions as possible, but do not duplicate any of it for commercial purposes. Thank you for your consideration.

        I am not a master accordion repairer/tuner. I have had no formal education or apprenticeship in repairing and tuning. I'm quite certain that I do not know everything there is to know about repairing and tuning accordions. I have learned some things about repairing and tuning by reading what others have written, but unfortunately, until now there has been very little available.

     I have learned primarily by experimenting, and that means often making mistakes and correcting them. I have made many mistakes and I have learned enough from correcting them that I am now quite comfortable doing most repairs and tuning dry and light musette (LMM and LMMH) accordions. However, I am just beginning to learn about tuning full musette (LMMM and LMMMH) accordions.

        Even though it is still very much a work in progress, I think writing down what I have learned so far may be of value to others taking a similar path. It also provides an opportunity for those who know more about tuning than I to critique what I have written, should they be so generous as to do so.

        The internet provides a wonderful forum for disseminating information, and for discussing and revising that information. It puts all our heads together, and provides unprecedented opportunities for learning from each other. I have learned much about repairing and tuning accordions by doing it. I learned some more by organizing what I have learned into what I hope is a cohesive, coherent, communicable format, presented here. Since publishing this website I have learned still more from several of my readers who have shared their knowledge and experience. I hope to continue learning by hearing other people's reactions to what I have written. If you see some way to improve or expand upon on what I have written, then I am eager to hear what you have to say.

George Bachich
April 2014


Reeds go out of tune for many reasons, and filing the reed is not always the best remedy.

        It is best to understand why each reed is out of tune before trying to correct its tuning by filing or scratching the reed tongue. Always look for clues that might indicate the cause of the problem and correct any such problem before tuning. Tune the reed only as a last resort. If you tune first, and correct the original problem later, the reed will once again be out of tune because of the filing or scratching you did.

        Dirt For instance, dirt and grime add weight to the reed tongue, throwing it out of tune. Dirt can also increase friction and drag. Cleaning a dirty reed will often put it back in tune, or at least closer to it, so whether or not you see any dirt, wax, or grime on the reed tongue, clean it thoroughly before attempting any tuning (but not before you establish your tuning bellows base reading and tuning bellows target for that reed). See the chapter on cleaning reeds and the chapter on using the tuning bellows.

        Rust - Rust on the reed tongue throws it out of tune by adding weight to it (iron oxide is heavier than iron). You should brush off all the rust using one of the small brass wire brushes described and shown below. However, after the rust is removed the reed will still be out of tune because it will now be lighter, due to the loss of the iron bound up in the iron oxide). Whenever rust is involved, cleaning, then filing the reed is the only way to get it back in tune. See the chapter on Putting File to Metal.

        Cracks - Metal fatigue cracks throw reeds out of tune, and filing will not help, at least not for long, because the crack will continue to develop, throwing the reed farther out of tune as it develops. Discard any cracked reed.

        Loose rivet - If the steel rivet (sometimes a screw) holding the reed tongue to the reed plate is not extremely tight, more of the reed tongue can lift off the reed plate as the tongue swings, effectively increasing the length of the tongue, thereby flattening the pitch. If a reed is difficult to tune because its frequency wanders more than that of its neighbors in response to changes in bellows pressure, try resetting the rivet before tuning.

        Misalignment – A loose rivet can also allow the reed tongue to drift out of perfect alignment with the vent. Misalignment of the reed tongue in the vent will usually silence the reed, but slight misalignment can sometimes allow it to keep working, while causing a metallic ringing noise as the tongue brushes lightly against the side of the vent on each oscillation. The friction caused by this metal on metal contact slows the reed significantly, throwing it out of tune. Very slight misalignment can cause the reed to go flat without making any ringing noise. The best remedy in all these cases is to correct the misalignment. See the chapter on Aligning Reed Tongues.

        Bad Leathers - A stiff, curled, loose, or missing leather can throw a reed out of tune by up to ten cents or more. Surprisingly, until you think it through, the reed tongue that is thrown out of tune is often the one alongside the defective leather, rather than the one under it. The best remedy is to replace the leather.

        New leathers -  A new leather anchored differently than the original (glued along more or less of its length than before) can also throw a reed out of tune (this time the one under it), by up to five cents or more. The easiest and surest way to correct this (once you have replaced the leather and rechecked the tuning in the accordion in order to determine the actual tuning error with the new leather) is to retune the reed with a scratcher or a file.

        Cracked wax - Cracked wax can throw a reed out of tune by allowing the reed plate to vibrate relative to the reed block (rather than transferring all its vibration energy into the reed block). This usually also causes a significant decrease in volume relative to neighboring reeds, and often creates a buzzing sound, but sometimes, in the early stages, the only symptom is the reed being out of tune on the flat side. If you scratch or file such a reed to tune it, your victory will be temporary at best, because the reed will continue vibrating itself loose, going farther out of tune until it finally goes silent and falls out. Then when you wax it back in, it will be out of tune because of all the filing you did. Especially if both tongues on the same reed plate are sounding flat, check for cracked wax by pressing the reed plate against the reed block with your thumbnail while the reed is sounding on your test bellows. If the finger pressure causes the reed pitch to sharpen, you will know that the wax has separated from the reed plate, allowing the reed plate to vibrate around the reed tongue. In this case, remove the reed from the reed block, clean all the old wax off the reed and off the reed block, and wax the reed back in with fresh reed wax. Then put the reed block back into the accordion and check the tuning again.

        A change in the weather - Changes in temperature, humidity, and barometric pressure all impact air density, which is one of the factors determining the frequency at which reed tongues vibrate. Changes in air density affect all the reeds in a similar manner, though, so if all the reeds were tuned under approximately the same atmospheric conditions, the accordion should at least remain in tune with itself as atmospheric conditions change. However, if the tuning job is not completed under the same atmospheric conditions under which it was begun, the reeds tuned under one set of conditions will not be exactly in tune with those tuned under different conditions. For instance, a 20 degree Fahrenheit change in temperature can change the tuning of a reed by 2 or 3 cents. 

        Microphones, plugs, etc. – Anything installed on top of a reed block that hangs over the edge impinges on the reed tongue’s environment. The reed responds to changes in its surroundings by changing its frequency. This can easily be demonstrated by hanging your thumb over the edge of a reed block right over a reed you are sounding on your test bellows. With your tuner turned on, you can observe the change in tuning as you move your thumb into and back out of the reed tongue’s immediate vicinity. Installation of microphone systems after the accordion has been tuned often result in one or more reeds going out of tune. You will have to decide whether to relocate any such obstacle or to tune the affected reeds to compensate for the obstacle’s presence.

        Register slides out of alignment - If the register slide is not in the fully open position, the reeds along that bank of reeds can be out of tune. In that case, moving the slide to the full open position will put them back in tune.

        Normal wear – Sometimes no reason can be found, in which case we conclude that the reed has gone out of tune as a result of normal wear. This is a cop-out, of course, because there is always a reason, perhaps as simple as the wax being harder than it was at the last tuning, but if we cannot find and correct any of the above problems, we have no choice but to tune the reed.

        Note that tuning by scratching and/or filing would be helpful only in three or four of the cases described above (rust, new leather, normal wear, and possibly microphones or plugs mounted to the reed block). In all the other cases, scratching or filing before solving the underlying problem would actually be counter-productive, so investigate before you file, and remedy any problems prior to tuning.

Spot Tuning

        Tuning a few problem reeds without re-tuning the whole accordion is called spot tuning. In this case, the amount of filing is usually minimal, if any, and most of the tuning is done with a scratcher. It is far less work than a total re-tuning, because the reeds are generally left in the reed block while you tune. If just a few reeds are out of tune by more than you can comfortably correct with scratching or diagonal filing, you can take them out of the reed block individually (after determining the magnitude and direction of the correction needed), tune them to within a few cents, then wax them back in for their final tuning. That’s a lot easier than re-waxing the entire reed block.


        Re-tuning, i.e., altering the basic tuning by tuning all the reeds to a different A, or changing a dry tuned accordion to wet (see the chapter on Tuning Musette Reeds), is a much bigger job than spot tuning, as many more reeds are involved and you will probably have to take the reeds out of the reed block for the first stage of tuning. Re-tuning requires a lot of filing in the first stage, and while mounted in the reed block, adjacent reeds and leathers get in your way, making filing difficult. If you have a lot of filing to do, it is much easier to file a reed the same way the reed maker does, with the reed plate out of the reed block and firmly clamped in a vise, with the leathers removed. See “Putting File to Metal” for more details on the mechanics of tuning.

Setting reasonable standards of accuracy

        Don’t be too obsessed with accuracy -
Whatever the nature of your tuning project, don’t waste time by tuning to greater accuracy than is needed. Electronic tuners are accurate to within 1/10 of a cent, but trying to tune all the reeds in an accordion to this degree of accuracy is a fool's errand for at least four reasons.

        First, most human ears cannot detect pitch differences smaller than about three cents, so especially on the lower notes there is little reason to exceed that level of accuracy.

        Second, the tuning of a reed varies from day to day by up to a few cents due to changes in temperature, atmospheric pressure, and humidity (all factors impacting air density). You can spend a lot of time meticulously tuning only to discover the next day (or even later the same day) that all the frequencies are off again.

        Third, some reeds can vary by a cent or more depending on how long they play. A reed may start out sharp and slightly flatten after sounding for a second or two, in which case there is no rational basis for choosing one condition over the other as a basis for your tuning (do you want your whole notes in perfect tune, or your quarter notes?).

        Fourth, the frequency of the reed varies with the volume (bellows pressure) because it flexes closer to the rivet as it makes wider swings in response to greater air flow. This brings more of its length into play, lowering its pitch. It is very difficult to tune every reed at exactly the same volume level, or even to check the tuning of the same reed at exactly the same volume twice in a row, so even if you meticulously tune with great precision, you will find it difficult to duplicate or verify your results, even just a minute later.

        For a partial solution to this problem, see the chapter on Tuning With a Manometer, but the fact remains that extreme accuracy is nearly impossible and not necessary, except in the case of the highest notes in your musette reed set. See the chapter on Tuning Musette Reeds.

        But don’t be careless, either, as there are other factors to consider which may make accuracy more desirable. 

        First, even though the instrument's absolute tuning will vary from day to day and at different volume levels, the various reeds in it will be influenced by the same external factors and will respond in very similar ways, thus keeping them more or less in tune with each other. So even though under certain conditions the accordion may be slightly out of tune with your electronic tuner or even with other instruments, it will be in tune with itself, which is the most important thing.

        Second, even though we cannot hear the difference between two pitches two cents apart played sequentially, we can hear the harmonics and possibly the beat created when they are played together, and this contributes significantly to the quality of the sound. Similarly, the natural harmonics and beats created by certain intervals and chords may vary enough to be noticeable to some people with well trained ears (especially at the higher pitches) even though the individual notes may not be perceptibly out of tune. At the lower pitches these harmonics are much less audible and much less sensitive to small tuning errors.

        Set reasonable standards of accuracy - For these reasons, if I am tuning an entire instrument, I tune dry reeds to within one cent at the highest pitches, to within three cents at the lowest pitches, and to within 2 cents near the middle of the accordion's range. If I am spot tuning, I tune to within one or two cents of the average of several neighboring reeds. After all, the whole accordion may be slightly out of pitch on the day I am spot tuning due to the factors listed above, in which case if I tune one reed to exact pitch as indicated by my electronic tuner, that reed may be out of tune with the rest of the instrument. It makes far more sense to check other reeds nearby, eliminate the outliers, and take an average of the rest. Then tune the subject reed and the outliers to within one or two cents of that average. 

        Accuracy is most important in musette tuning - Note that the above standards apply only to dry reeds. You will find that when tuning musette reeds, a tuning error of a couple of tenths of a cent can make a noticeable difference at the higher pitches. However, in that case you are usually tuning to the beat, rather than to an absolute note value. When tuning to the beat, all that matters is the difference in pitch between the two reeds. This difference can be controlled with great precision because both reeds are sounding together, are operating under identical atmospheric conditions, and are responding to the exact same bellows pressure. Even if atmospheric conditions and/or bellows pressure are different next time, causing both notes to be sharper or flatter, they will both be impacted in approximately the same way. The difference between their frequencies will remain pretty stable, and so will the beat, which is all that matters when you tune to the beat. See the chapter on Tuning Musette Reeds.


        While it is true that tuning accordion reeds involves removing metal from the reed tongues (see the chapter on Putting File to Metal), the real essence of tuning lies in investigation, analysis, and planning. Whether tuning an entire accordion or merely spot tuning a few problem notes, before putting file to metal it is absolutely essential to develop and adhere to a tuning plan that tells you how much to change the tuning of each reed tongue and in which direction.


        A=440? - The first step of your investigation is to learn or decide whether this is or will be an A440 accordion or whether it is or will be tuned to some other value for A. An A440 accordion has the A above middle C (the second A up in the clarinet reed set) tuned to 440 hertz, and all other notes follow from that. You will find accordions tuned to A440, A441, and A442, as well as other values. I found one 1933 model tuned to A446.

        The easiest way to determine which value of A forms the basis of your accordion’s original tuning plan is to use a strobe tuner or "virtual" strobe tuner to examine a few of its bassoon reeds or bass reeds to find which A setting gives you an indication of zero cents deviation, or at least the smallest average deviation. Naturally, this investigation must be done with the reeds still mounted in the accordion, because they will vibrate at different frequencies once you take them out of the accordion. You should choose to investigate the bass reeds or the bassoon (L) reeds because they are always straight tuned, whereas one or more sets of your clarinet (M) reeds may be wet tuned, which would give you a misleading result. Piccolo (H) reeds, can also be checked, but being smaller, they are more vulnerable to being out of tune due to an accumulation of dirt and grime, which could also give you a misleading or ambiguous result.

        You can tune your accordion to any A you want, but discovering and adhering to the original basic tuning is a much easier task than switching to a different value of A. Adhering to the original tuning plan will require much less filing on fewer reeds, and you will probably be able to do it without removing the reeds from the reed blocks.

        Wet? Or dry?- The second step of your investigation is to determine whether the accordion is wet or dry tuned. That is, does it have more than one set of M reeds, and if so, is one or more of those sets tuned slightly off-pitch to create a tremolo effect (wet tuning, often called musette tuning) when both (or all three) sets play? (For clear instructions on how to identify your accordion’s reed configuration, see Chapters 4, 5, and 8 of PIANO ACCORDION OWNER’S MANUAL AND BUYER’S GUIDE, available on the "Owner's Manual" page).

        If you have two M sets, to determine if it is wet tuned you simply listen for the tremolo. If your ear is not already trained to recognize tremolo, then set your treble switches to play just one M set. Play a single note and while sustaining that note hit the switch that allows just one other M set to play that note along with the first. If you suddenly hear tremolo, it means one of those reeds is slightly out of tune. If you hear tremolo on nearly all notes when both M reeds for those notes sound, the detuning is probably intentional, meaning you probably have Musette tuning. If you hear tremolo on just a few, you probably have a dry tuned accordion in which those few reeds are out of tune. If you hear tremolo when just one clarinet reed is played along with a corresponding bassoon or piccolo reed, then one of those reeds is definitely out of tune.

        If you have three M reed sets, then you certainly have musette tuning. Use your strobe tuner to identify which set is supposed to be on concert pitch (I call that the clarinet reed set), which set is tuned sharp, and which set is tuned flat (I call these detuned reeds the musette reed sets). In some LMMM and LMMMH accordions both musette sets are tuned sharp, with one being sharper than the other, in which case you will discover this during your investigation. If you have an A440 accordion and an iPhone or an iPad, you can more conveniently investigate all this by using the "Strobe Tuner Pro" app from Marcus Cavanaugh, inc. ($2.99 at the iPad App Store).

        Once you have identified which reed sets are which, label the reed blocks accordingly for future reference. I label each row on each reed block: “bassoon white”, “bassoon black”, “clarinet white”, “clarinet black”, “musette white”, “musette black”, etc. (remember, each reed set is divided approximately in half, with each half mounted on a different reed block).

Analysis and planning 

        If you have a dry tuned accordion and wish it to remain dry, there is relatively little analysis and planning to do. You will simply plan to tune all reeds including bass reeds to concert pitch, based on the A (440, 442, etc.) of your choice. If you adhere to the original basic tuning, you will probably not tune every reed, but rather just those that are out of tune by some minimum amount, such as 2 cents or 3 cents or more, especially for the lower notes. See the chapter on tuning dry reeds.

        If you have a musette accordion (or if you are converting your dry tuned LMMH accordion to musette, the first step in your tuning plan will be to tune all the dry reeds (the bass, bassoon, clarinet, and piccolo reed sets) to concert pitch (based on the proper A for your accordion). But the second and more interesting step is to develop a tuning plan for your musette reeds. Before tuning any musette reeds, you must first arrive at a musette tuning plan that achieves the tremolo beat progression you want to create.

        If you are merely spot tuning to correct a few bad sounding notes, then you will want to analyze and understand your accordion’s original musette tuning plan (beat progression), so you can tune those few problem reeds to fit the current plan. But if you are tuning the whole instrument, your goal may be to alter its tuning plan/beat progression to one that sounds better to you. 

        Many possible tuning plans - For LMM and LMMH accordions, establishing a tuning plan begins with deciding what beat progression you want. For instance, do you want to hear one beat per second (bps) on the lowest F, with the beat frequency gradually and steadily increasing as you ascend the 41 note harmonic scale to four bps on the highest A? Do you prefer to hear two bps on the lowest F, progressing to eight bps at the highest A? Maybe one bps progressing to six bps is more to your liking, or two to four, or some other progression. Although my American ear does not appreciate tunings wetter than these, in some parts of the world some people claim to like tunings as wild as 4.5 bps at the low F and as much as 30 bps at the high A.


        Thierry Benetoux, in THE INS AND OUTS OF THE ACCORDION, briefly describes four different beat progression curves, running four full octaves from the D# below our lowest F (which I will call F1) to the D# above our highest A (which I will call A4 (because it is the fourth A on our 41 key keyboard). Mnsr. Benetoux describes his vibrato beats in terms of beats per minute, but if we divide his numbers by 60 to convert to beats per second, and if we plot out his curves, then truncate them to fit our 41 key piano accordion keyboards, we come up with these four curves:

    "American att." runs from 0.7 bps at F1 to 3.5 bps at A4.

    "American" runs from 0.9 bps at F1 to 4.2 bps at A4.

    "Light musette" runs from 1.6 bps at F1 to 5.3 bps at A4.

    "Musette" runs from 2.2 bps at F1 to 7 bps at A4.

        Although Mnsr. Benetoux has drawn simple straight lines to represent each of his curves, none of his curves are actually simple straight lines. If you plot them on graph paper with the 41 key scale across the bottom, you can readily see that they all have an inflection point somewhere in the first half of the keyboard, after which the curve rises more steeply.

        One way to come a bit closer to these curves is to create a linear progression of cents deviation between successive notes, rather than a linear progression of beat frequency. That is, make the incremental decrease in cents deviation between any two adjacent notes be constant, rather than making the incremental increase in beat frequency be constant. The result is very similar, except that a little less of the increase in beat frequency occurs in the lower half of the keyboard and a little more of it occurs farther up the scale. The increase in beat frequency as you advance pitch-wise up the keyboard is still very smooth and the beat frequency increase between any two adjacent notes is still barely perceptible, if at all, but the overall sound of the accordion will be slightly different because the beats will all be a bit slower near the middle of the keyboard than they would have been if you had used a linear beat progression.

        For LMMM and LMMMH accordions (called “full musette” accordions), you must not only decide on a desired beat progression for each of the two musette sets paired individually with the dry clarinet set, but you must also consider the beat progression that will result when all three are played together. I don’t yet know how to predict that beat progression, so I can offer you little guidance on this, other than to suggest that you analyze the beat progressions of the individual pairs in a similar accordion whose tuning you like and duplicate that in your own accordion.

        The main point is that there are a very large number of different tuning curves (beat progressions) possible, and the one that is right for you depends on your personal taste. Whatever your choice, you must then establish the desired beat frequency for each note in between. You might want a nice linear progression, with the beat increasing by the exact same amount between any two successive notes on the scale, or you might want to increase the beat slowly through the first octave or so, then increase it more rapidly the rest of the way up the keyboard. Another possibility is to increase the beat to some maximum near the middle of the keyboard and keep the beat frequency constant from there on up. The only hard and fast rule is to avoid abrupt changes. No note should beat at a significantly different frequency from its immediate neighbors. If you are not yet sure what you want, then before tuning your accordion to a new beat progression, you should listen to many accordions, analyze the beat progression of one you particularly like, and duplicate that tuning on your own accordion.

        For a quick comparison of several different beat progressions, visit the Fisitalia website and play the short recordings of the six different tunings they offer. Unfortunately, they don't precisely define the  full beat progression of each tuning, and the notes played for comparison are only through a one octave C chord near the low frequency end of the keyboard, but you can easily tell the difference, learn to count the beats, see for yourself how the beat frequency increases as you advance up the scale, and perhaps decide which tuning sounds best to you. However, keep in mind that their recordings do not reveal the tunings at the high end of the keyboard, where the beats would be faster, and where really fast beats can sound really awful. Here is the link.

       Analyzing beat progressions - One way to analyze a beat progression is to compare the beat on each note to a metronome beat. Several digital metronomes are available as free downloads. I use an iPad app called Epic Metronome. I like it because it has a moving graphic presentation of an old fashioned mechanical metronome arm swinging back and forth across the screen, and that visual representation is easy for me to monitor while I’m counting the musette beat. I set the metronome for 30 beats per minute, then I count the beat during two full swings of the arm, which is four seconds. I divide that count by four to get the beat frequency to the nearest quarter beat per second (bps).

        If you have not yet trained your ear to count the beats, you might start by singing silently along with them, la-la-la-la-la-la, matching each "la" to one of the beats. For faster beats, it helps to think in terms of pairs, triplets, and quadruplets of "la". For instance, while you might not be able to count ten individual "la"s per second, you might be able to count two quadruplets and a pair. With this technique, I can count with confidence up to about 12 bps, which to my taste is beyond the desirable range.


        By counting the tremolo beat for each note, you can accurately plot the tuning curve for any light musette accordion (LMM or LMMH). As you plot the beats for each note, you are likely to find some deviations from a smooth curve. If you assume these deviations are caused by specific notes being a little out of tune, you can ignore them in order to plot a smooth curve (or a straight line) as a model for tuning your other accordions.

        However, for some of the wider tunings that you might encounter in some full musette (MMM, LMMM or LMMMH) accordions, you might not be able to count the beat near the upper end of the keyboard range, because it is too fast. In this case, unless you can hear the beat in terms of its pitch as an overtone (can anyone do this? I don't know!), you will have to revert to tuning those notes entirely in terms of cents.

        BPS versus cents - Although I think it is most informative to speak of musette tuning (at least light musette tuning) in terms of beat, most people still speak of it in terms of cents. For instance, I once saw a list that showed American musette as 10 cents sharp, Italian musette as 15 cents sharp, French musette as 20 cents sharp, Irish as 25 cents sharp, etc. I have seen some sources suggest tunings as wild as 32 cents sharp.

        Aside from the fact that these wider tunings would sound awful to me, I think their descriptions are misleading, because they seem to suggest that the entire range of notes should all be tuned to the same cents deviation, which is never the case. Sometimes the given number refers to the deviation at the low F on our keyboards, and sometimes it refers to the deviation at A=440. But it never refers to the deviation at the highest A.

        Regardless of the cents of deviation used at the low end of the keyboard, the deviation is always reduced as we progress up the scale. If it were not, few of us would like the sound. Tuning the high end or even the middle of the keyboard of an LMM or LMMH accordion to large deviations creates a beat so fast that to most American ears it sounds more like fingernails on a blackboard than a beat, and it definitely grates on our nerves. (In MMM, LMMM, and LMMMH accordions, it is said that the presence of the third reed makes the wider tunings tolerable. I plan to do some experimenting on this in the near future as I develop tuning plans for my LMMM acordions.)

        Regardless of the amount of deviation, or of the differences among our individual tastes, we don’t hear the cents, we hear the beat. The beat frequency is extremely important to the quality of the sound. When tuning musette reeds, we most often tune them to the beat, rather than to some nominal number of cents sharp or flat. Our tuning plan is based on a desired beat progression. What we hear and what we really seek to control is the tremolo beat frequency measured in beats per second. It therefore makes much more sense to describe musette tuning (at least light musette tuning) in terms of beat, rather than cents, and to specify how the beat changes as we progress up the keyboard. 

        Understanding the relationship between hertz, bps and cents - The problem with this approach is that our tuning equipment is calibrated in cents, (one cent equals 1/100 of a semitone), so to be able to use our tuners to achieve a desired beat, we must also understand the exact relationship between cents and beats per second. Unfortunately, that relationship is different for each note. Generally, one cent of difference in pitch has less impact on the beat frequency at the lower notes and more impact at the higher notes. This is because the beat frequency is exactly equal to the difference between the two note frequencies expressed in hertz, and because each cent represents more hertz at the higher notes.

        For instance, at the lowest clarinet F on our keyboard (174.6 hertz), each hertz of difference (bps) equates to about 10 cents. Near A=220 hertz, each hertz equates to about 8 cents, at A=440 hertz, about 4 cents, at A=880 hertz, about 2 cents, and at the highest clarinet A on our keyboards (A=1760 hertz), each hertz is only about 1 cent (for a more complete discussion of why this is so, see Chapter 27 of PIANO ACCORDION OWNER’S MANUAL AND BUYER’S GUIDE). The mathematical formula for determining the number of cents equivalent to one beat per second at each note boils down to the following:

        For any given note whose frequency in hertz (cycles per second) is F, the frequency of a note one cent higher is F x 2 to the 1/1200 power, the frequency of a note 2 cents higher is F x 2 to the 2/1200 power, and so on, right on up to a note 100 cents higher, i.e., a full semitone higher, having a frequency of F x 2 to the 100/1200 power (the same as the 1/12 power of 2, which is the basis for an equal temperament 12-note scale). 

        Musette tuning calculations - (We won’t actually do much calculation here, but if even the passing discussion of math bores you, skip now to the cheat sheet.)

        For a convenient example, let’s take the case of a note 10 cents higher than frequency F. According to the formula, this higher note has a frequency of F x 2 to the 10/1200 power. The hertz difference between these two note frequencies is F x 2 to the 10/1200 power minus F. This hertz difference is exactly equal to the beat frequency that will be generated by playing these two notes together (for a clear, plain language explanation of why this is so, see Chapter 27 of PIANO ACCORDION OWNER’S MANUAL AND BUYER’S GUIDE). Since we are comparing two notes exactly ten cents apart, this beat frequency is the number of tremolo bps that result from ten cents deviation at this particular note. To get the number of cents deviation required to generate just one tremolo beat per second, we must divide 10 by this number. 

        For instance, applying this formula (F x 2 to the 10/1200 power) at piano A4 (440 hertz), tells us that a note ten cents higher has the frequency 442.55 hertz. Subtracting F (440) from 442.55 gives us the difference between these two notes, which is 2.55 hertz. Since each hertz of difference generates one beat per second, we know that these two notes played together will generate a beat of 2.55 bps. In other words, ten cents of difference generates a beat of 2.55 bps at A440. Dividing 10 by 2.55 tells us that 3.92 cents difference generates one beat per second at A440.

        The cheat sheet - This formula may appear difficult to evaluate, but in fact it is quite easy on an Excel spreadsheet, and I have done it for you and posted the results below. The chart shows how many cents difference are required at each note to achieve 1 bps tremolo beat. Double that many cents will generate a 2 bps tremolo beat. If you want to generate a 3.25 bps tremolo beat on a particular note, multiply the cents value of 1 bps by 3.25.


        This chart is an essential aid when we are tuning to the beat. Let's say we hear a beat of 2.25 bps on one note that, according to our tuning plan is supposed to produce a beat of 3.25 bps, that is, it should beat 1 bps faster. Assuming this particular musette set is tuned sharp, the chart tells us how many cents we need to further sharpen this particular reed in order to increase the beat by one more bps. If we are tuning a set that is tuned flat, it tells us how many cents we need to flatten it. If the beat of a particular note needs to be raised or lowered by 0.75 bps, then multiply the number in the chart by 0.75, and so on.

        This is your cheat sheet. It allows you to quickly calculate how many cents correction you need to apply to the current tuning in order to alter the current beat pattern to fit your tuning plan for each individual note. It is the single most essential ingredient for successful musette tuning, yet I have never found it published anywhere. As far as I know, this is its first appearance.


        With this background information in mind, we can now recap in outline form the proper procedure for tuning a musette accordion, which is the most complex type to tune. Details for steps 1, 2, and 3 have already been provided. Details for steps 4, 5, and 6 are provided later on. For a dry tuned accordion, skip steps 5 and 6.

    Step 1: Determine the basis of its tuning (A=440, etc.). Set your electronic tuner to this standard, and save the setting (or the next time you turn it on, it may revert to the default setting).

    Step 2: Check all reeds for security and condition (wax, leathers, rust, grime), and remedy any defects.

    Step 3. Identify and label each reed set.

    Step 4: Tune dry reeds as necessary

    Step 5: Develop your musette tuning plan in terms of the beat progression you want.

    Step 6: Tune your musette reeds, one musette reed set at a time.


Step 4: Tuning dry reeds

        Using the tuning bellows - When tuning, it is important to understand that each reed tongue will vibrate at a different frequency in the accordion than on your tuning bellows. This is due to several factors including reversed air flow on the tuning bellows and differences in physical surroundings. Although the reeds will have to be tuned outside the accordion on your tuning bellows, you will want them to be in tune when installed in the accordion. You must establish a tuning bellows target value for each reed tongue that will put it in tune once it is back in the accordion.


        For your dry tuned reeds, measure how many cents each reed tongue is out of tune, and in which direction, while the reeds are installed in the accordion. Unless you are using Dirk's Tuner, which can listen to more than one reed at a time, you must turn off all reed sets other than the one you are measuring.

        If you have an A440 accordion, you can use the Marcus Cavanaugh iPad/IPhone app referenced above, but if you are tuning to some other value of A, then you will have to use a strobe tuner or virtual strobe tuner that can be set to that standard. I use the Peterson VS-II Virtual Strobe tuner, which used to list for $359, but has been discontinued and is now available used in places like eBay, often for around $100. Express any deviation you find as a correction factor. For instance if you find that a reed is three cents flat, then your correction factor will be to sharpen it three cents. Use the tuning bellows to measure and determine when you have achieved your correction factor.


        For the sake of efficiency (to avoid opening the accordion hundreds of times) measure the tuning of all the reeds in one set (bassoon, clarinet, or piccolo) at one sitting and make a chart showing the amount and direction of the required correction for each reed tongue. Then test those same reeds on the tuning bellows (still mounted in their reed blocks, naturally, and with leathers still installed), and note on the chart how far out of tune each one appears to be on the tuning bellows and in which direction. You will not be tuning to eliminate this indicated error, as it will probably be different from the real error noted while the reed block was in the accordion. You will use this tuning bellows reading only as a base line for calculating your tuning bellows target value for each reed tongue.

        Calculate the desired tuning bellows target value by applying the correction factor determined while the reed block was still mounted in the accordion, and note the tuning bellows target value on your chart. Then tune the reed tongue to that target value on the tuning bellows. Use the tuning bellows just to measure the amount of correction being made. The tuning bellows target value will not be zero (zero cents out of tune) unless the reed happens to have the same frequency inside the accordion as it does on the tuning bellows, which is uncommon.

        For example, a typical reed tongue might be out of tune by +6 cents (6 cents too sharp) in the accordion. In that case you will note a required correction factor of minus (-) 6 cents (flatten it by 6 cents to bring it to zero cents out of tune in the accordion). If that same reed tongue is out of tune by +9 cents on the tuning bellows, your tuning bellows target value will be +3 cents (9 minus 6) and you will flatten it by 6 cents to get it there. When you put it back in the accordion and test it again, it should be exactly in tune. 

        As a final check after tuning to make sure your allowed dry reed tuning tolerances are not too great, play each possible dry pair of reeds (a dry pair being the one you are tuning plus its octave mate from another dry reed set) together and listen for any beat that would indicate one of the reeds is out of tune, and retune any problem pairs to eliminate the beat. For instance, if your tolerances allow the low F in your bassoon set to be up to 3 cents flat and the low F in your piccolo set to be up to 2 cents sharp, and if both happen to be at opposite extremes of their tolerance ranges, this 5 cent difference might result in an audible beat when they are played together. In this case, one of those reeds will have to be tuned closer to the value of the other to eliminate the beat. In most cases with these stated tolerances, the pairs won’t be at opposite ends of their allowable ranges and will be well enough matched to preclude any noticeable beat. Only if one of the reeds is near one limit of the tolerance and the other reed is near the opposite limit will you have a problem.

        Have a clear, well organized worksheet where you can post the current tuning bellows reading, the correction factor to be applied, and the resulting tuning bellows target value for each reed tongue. It helps to color code the worksheet columns containing your inner reed tongue readings, correction factors, and targets to help prevent inadvertently reading the value for an external tongue while trying to tune an internal tongue. Constantly double check to ensure that you are looking at the correct target value for the reed tongue you are tuning and that you are tuning the correct tongue on the correct reed in the correct reed set, and that you are tuning it in the right direction (i.e., not inadvertently sharpening it when you actually need to flatten it).


        Organization, focus, and perspective- Skill, dexterity, and judgment are all required to successfully tune, but of equal importance are organization, focus, and perspective. It is important to be organized and focused, yet always keep in mind the big picture in order to avoid inadvertently tuning the wrong reed, or the wrong tongue, or tuning in the wrong direction, or by the wrong amount.

        Many pitfalls - There are many opportunities for error, confusion, or distractions to ruin your tuning job. For instance, your tuner could be set to A=442 or some other value not applicable to your accordion, which would give you the wrong readings for every reed. Or you could misread the tuner readout and write down that the reed must be sharpened, when it really needs to be flattened. You could accidentally enter the value for the inner reed tongue in the column for outer reed tongue values, or on the line for a different note, or accidentally write a minus sign when you meant to write a plus sign, or misread your own handwriting. Once you have the values written down, you could miscalculate the correction required or you might calculate correctly but use a wrong value in the calculation, perhaps due to reading it from the wrong column on your chart. 

        Even if you avoid all measurement, calculation, and tabulation errors and produce a perfectly accurate chart of all the reeds and their required corrections and tuning bellows target values, there are still plenty of opportunities for error during the actual tuning of the reeds. You might inadvertently tune the inner tongue to the value desired for the outer tongue, or tune the wrong reed altogether, or be on the wrong reed block, or sharpen a reed when you should have flattened it. You might be closing the tuning bellows (sounding the interior reed) while thinking you are checking the tuning you just did on the exterior reed. Whatever the reason for the error, I know from experience that it is very disheartening to find that you have taken metal off the wrong reed tongue, or off the wrong end of the right reed tongue.

        You should never take your eye or hand off the reed you are tuning. If you must look away to pick up a tool, then keep your finger on the reed plate you wish to tune. If you don't, you may find yourself tuning the wrong reed. Practice on accordions of little value in order to learn to systematically avoid all these pitfalls before tuning an accordion of value.

        Even if you have got everything right so far, and are tuning the correct reed tongue in the right direction toward the correct target value, you are still not out of the woods. It is very easy to accidentally take off too much metal and overshoot your tuning target. Most often this happens when light filing or scratching is apparently not changing the tuning (or maybe even changing it in the wrong direction) so you get a little more aggressive with it.

        This frequently occurs when your scratching or filing is displacing (relocating) metal, but not actually removing it, or perhaps displacing enough metal to cancel out the tuning impact of any metal you remove. For instance, filing the tip of the reed tongue (to sharpen the pitch) can actually flatten the pitch by shifting burrs of metal (mass or weight) toward the reed tip rather than removing them entirely. Similarly, scratching the lower third of the reed (to flatten the pitch) can sometimes sharpen the pitch by plowing up tiny burrs and shifting them toward the fixed end of the reed rather than removing them entirely. If the next pass of the file or scratcher finally removes these burrs, you will get a sudden jump in the reed frequency, perhaps well beyond your target value. Then you are faced with taking metal off the opposite end to correct the overshoot. You can avoid this trap by carefully and thoroughly scraping off all burrs before taking each measurement.

        In any case, the important thing to remember is that you sharpen pitch by shifting the reed tongue's center of gravity toward the fixed end, and you flatten pitch by shifting the center of gravity toward the free end. You can shift the center of gravity either by removing metal or by shifting metal around. For more details on how this is done, see the chapter on Putting File to Metal. 

        There are many other tuning pitfalls to avoid. If you let your thumb or any part of your hand hang over the edge of the reed block above the reed while sounding that reed on your tuning bellows, the proximity of your hand will flatten the pitch of the reed. This effect may overshadow the effect of your most recent scratching or filing, providing erroneous feedback and leading you to erroneous conclusions about whether or not additional filing is required. Not having the opening in the bottom of the reed block fully over the opening in the top of your tuning bellows will also give erroneous readings. Gripping the reed block tightly while filing or scratching a reed can compress or displace or wrinkle the leather on some other reed, which can change that other reed's tuning. Similarly, replacing a leather can change the tuning, especially if the length of the glued-down portion of the new leather is not exactly the same as on the old. When you begin tuning that reed, blindly applying the previously determined correction factor may produce an undesirable result. If rough handling or replacing leathers has changed the tuning of any reed, a new correction factor must be determined for that reed before attempting to tune it. Naturally this requires putting the reed block back in the accordion and checking the tuning of that reed again.

        Great care must be taken not to damage the reed tongue or the reed plate while tuning. This is especially important in the highest pitch reeds in the piccolo reed set, say, from the highest A down to the C below. It is very common to find these reeds silent due to being damaged by excessive force during a previous tuning. I have found them arched, cupped, twisted, even kinked. Reeds with any of these defects are impossible to voice properly, and should be replaced with new. However, they can sometimes be straightened with a bit of tedious careful work under 10x - 20x magnification (see the chapter on “Straightening Piccolo Reeds”).

        Reeds can be ruined by removing too much metal in one place. Creating thin spots can theoretically change the timbre and encourage cracks. Reeds can also be ruined by bending them too far while getting them in position to file, or by pressing too hard with the file or the scratcher, or by filing or scratching without providing proper support behind the reed tongue (the smaller the reed, the more vulnerable it is to this kind of mishandling). Inadvertently narrowing the reed tongue or shortening its tip can ruin a reed by increasing the clearance between the reed tongue and the vent. Scratching or nicking the reed plate at the edge of the vent can similarly ruin the reed. 

        Occasionally the file will drag a metal burr or metal filing to the edge of the reed tongue, where it will wedge between the reed tongue and the vent, preventing the reed from sounding properly. These can usually be removed by passing a .0015" metal feeler gauge between the tongue and the vent, or by dragging it along the edge of the reed tongue, or sometimes by plucking the reed until it vibrates freely. However, don't push the feeler gauge too far into the vent, or you may damage or dislodge the leather on the opposite side. And don't pull the reed too far up to pluck it, or you may alter its voicing.

        More insidious is the effect of a tiny burr on the edge of the tongue that serves to slow the reed just enough to put it slightly out of tune rather than stopping it entirely. This can mislead you into thinking the reed must be sharpened in pitch, when all it really needs is to be wiped clean. If you file it again before wiping it clean, you may overshoot your target value. Similarly, the file or scraper or your finger can drag soft wax or other contaminant to the reed tongue and deposit it there, which will alter the center of gravity and thus the frequency until you wipe it clean. It is important to remove all wax from the top of the reed plate before tuning any reed, in order to minimize the possibility of dragging that wax onto the reed tongue during tuning.

        Occasionally, your efforts to support the interior reed tongue from below will dislodge or damage the interior leather, throwing the interior reed tongue alongside of it suddenly out of tune and skewing your readings for it. If you don’t notice this problem right away, the erroneous and misleading feedback from your electronic tuner may cause you to file or scratch more than necessary. Once you discover and correct the problem, you may find that you have significantly overshot your tuning target and that you now have to tune that same reed tongue back in the opposite direction.

        One major difficulty in checking the tuning and determining correction factors is the fact that reed frequency decreases (flattens) under higher air pressures, especially on the lowest frequency reeds. You might have a reed perfectly in tune at one bellows pressure, yet the next time you test it you will get a different reading if you use a slightly different bellows pressure. This makes it impossible to tune efficiently, as you cannot tell whether the change in reading is due to the scratching you just did, or to differences in bellows pressures. And if you have one reed perfectly in tune at one bellows pressure and another reed perfectly in tune at some other bellows pressure, then whenever you play those notes together, they will be out of tune with each other.

        Tuning at constant pressure with a manometer - This problem can be avoided only by tuning and testing at least all the low note reeds at the same bellows pressure (the low notes are the most sensitive to pressure changes). A simple home-made U- tube manometer can be used to measure bellows pressure and to provide a reference that allows you to apply the same bellows pressure to every reed. I used some old clear plastic medical oxygen hose for the tubing, food coloring to color an ounce or so of water, an eye dropper to put the colored water into the clear tubing. I mounted the tubing on an empty white plastic gallon jug to hold it in the proper U-shape, to support it in an upright viewing position, and to provide a white background behind the red liquid.



        To work properly, one end of the tubing must be open to the atmosphere and extended a few inches above the liquid level to prevent any liquid from being blown out under bellows pressure.  The other end must be looped to a similar height and connected to the bellows through some convenient length.

        To connect the manometer to the tuning bellows I drilled an oversized hole in the top cover plate and used a piece of foam rubber as a stopper to seal around the hose. Before installing it, I made a hole through the foam stopper and forced the hose through the hole.

        To connect the manometer to the accordion I remove the bass plate and the air release button and valve, and insert the manometer hose into the air release valve port, again using a piece of foam rubber as a stopper to seal around the hose. I route the hose out through the air release button hole in the bass plate and reinstall the bass plate so I can play the accordion while it is hooked to the manometer.

        I mark the level of the liquid at rest in the manometer on the plastic jug behind each leg of the U-tube with a felt pen, then while playing the accordion normally, I note the fluctuations in liquid level with varying bellows pressure and sound volume. Based on those observations I select an appropriate liquid level (bellows pressure) to use in tuning and mark that on the jug behind each tube, one for each bellows direction. When measuring the tuning of any reed in the accordion or on the tuning bellows, I can now hold constant bellows pressure at the appropriate level using the manometer as a reference. 

        Atmospheric factors - Even at constant bellows pressure, the tuning of any reed will vary from day to day due to changes in air temperature, humidity, and barometric pressure, all of which impact air density, one of the important determinants of the frequency at which any reed will vibrate. Expansion due to temperature changes also minutely changes the length of the reed tongue and hence the location of its center of gravity relative to the fixed end, another major determinant of the reed's vibration frequency.

        We intuitively expect to see reeds slow down due to an increase in temperature, as the reed tongue expands, carrying its center of gravity away from the fixed end. However, it is apparently not quite that simple, as on a recent test of three reeds (Piano A3, A4, and A5) at three distinct temperatures (48, 68, and 88 degrees Fahrenheit) I found, contrary to my expectations, that all three reeds were actually 1 ½ to 2 ½ cents sharper at 68 degrees than they had been at 48 degrees, although at 88 degrees the trend had reversed and the largest one was back to being about half a cent flatter than it had been at 68 degrees. (I would need more data points to plot a curve that would reveal at what temperature the trend actually reverses for each reed.) I can offer no theoretical explanation of why this reversal occurs.

        Tuning variations due to temperature changes are not a problem if all the reeds are tuned at about the same temperature, because they will change together. But if you tune one reed set on a 60 degree day and another on an 80 degree day, you may have to compensate for the temperature difference.

        The lowest notes often seem unstable - The lowest note bassoon reeds are more difficult to tune because they tend to drift in frequency, even at constant bellows pressure. The higher notes vibrate at a consistent frequency that your tuner can reliably measure to within one tenth of one cent, but on the lowest notes the tuner readout fluctuates, sometimes by several cents, sometimes eventually settling down to one value, but more often oscillating through some range of values. After you file or scratch, you can't know for sure how much of the change in readout is due to your tuning and how much is due to drift, or to measuring at a different phase of the oscillation cycle. You have to take an average, or decide to always use the value showing after some arbitrary but consistently measured period of time, such as two seconds after the reed first sounds. I hypothesize that this drift may be due to the uneven weight distribution along the length of the reed tongue (the largest treble reeds are much thicker and heavier at the tip), or it may be due to the large mass of the reed tip intermittently mobilizing more of the reed tongue’s length by occasionally lifting it off the reed plate closer to the rivet. Whatever the reason, you will not be able to tune these reeds to the degree of accuracy that is possible at the higher notes.

        Final check - Once all the reeds in a set have been tuned, check the operation of every reed in both directions on your tuning bellows (in case any have become jammed by tuning debris) then put the reed blocks back in the accordion and check the tuning again. This will highlight any errors or discrepancies that you will need to correct in a second round of tuning. Also check for proper voicing, as the forces imposed on the reeds during tuning can upset the voicing (especially on the smallest reeds), and you will want to correct any voicing problems at the same time you are correcting any final tuning discrepancies.

        Voicing the reeds - Voicing involves setting the free tip of the reed tongue to the optimum height above the reed plate. The optimum height is that which will require the least possible amount of air to get the reed tongue moving, yet not so close that it chokes (balks under sudden high bellows pressure). The proper height is typically about one reed thickness, but this is really just a starting point. The actual optimum height must be determined by feel and by ear. If the clearance is too great, the reed responds slowly and sluggishly, especially under light bellows pressure, because too much air leaks through the resulting excessive clearance around the reed tongue. If the clearance is too narrow, the reed will not sound at all under sudden high bellows pressure, presumably because its initial stroke is too short to accumulate enough spring energy to overcome the air pressure and bounce back, as would be required for it to begin oscillating. Finding the sweet spot can be tricky because if the clearance is ever so slightly less than the optimum, the reed can partially choke, that is, it can refuse to sound immediately, but eventually get going, which can sound a lot like a reed responding sluggishly because it has too much clearance. This can fool you into reducing the clearance some more, which will cause it to choke completely on the next trial. Generally, if a reed's threshold of response under very light bellows pressure is identical to its immediate neighbors, it is probably properly voiced. Any reed that requires more bellows pressure than its neighbors probably needs voicing. For detailed instructions on voicing reeds, see the article on page "Accordion Repair 1".

Step 5: Developing your musette tuning plan

        Developing your musette tuning plan involves several tasks:

    a) deciding what beat progression you want for each musette reed set,

    b) documenting the current beat progression for each M set paired individually with the dry clarinet set,

    c) comparing the existing beat progression to your desired beat progression to get a tuning correction factor measured in bps for each reed tongue,

    d) calculating an equivalent tuning correction factor measured in cents for each reed tongue,

    e) measuring each reed tongue’s current tuning on the test bellows in terms of cents, and

    f) applying the calculated cents correction factor to get a tuning bellows target value (in cents) for each reed tongue.

        I usually complete steps a) through d) for the whole reed set before going on to e) and f), which I complete for each reed tongue just before I clean and then tune it. I tune each reed tongue to match its tuning bellows target value before going on to e) and f) for the next reed tongue.

        Here is a more detailed explanation of each task:

    a) Deciding what beat progression you want was covered earlier under “Many possible tuning plans”. To be useful during tuning, it must be in the form of a list of the desired beat frequencies to the nearest 0.1 bps for each note on your treble keyboard. 

    b) Documenting the current beat progression was partially covered earlier under “Analyzing beat progressions”. To that I can add a few tips: With the reeds in the accordion, and with the treble switches set to play just the dry clarinet set and the single musette set you are tuning, play each note in both bellows directions and make a list of the current beats for each note in each bellows direction. If switches don’t allow this reed combination, then insert paper strips under the reed blocks to block air flow to the rows of reeds you do not wish to hear. If your accordion has more than one musette reed set, repeat this for each of those sets individually paired with the dry clarinet set.

    c) Comparing the current beat progression to the desired beat progression involves simple subtraction and listing the correction required (in bps) in the proper column for each reed tongue, paying careful attention to the direction of the required correction.

    d) Calculating an equivalent tuning correction factor measured in cents is just multiplying the bps correction factor by the cheat sheet bps-to-cents conversion factor for that note. 

        I use a chart which I call my Musette Reed Tuning Worksheet. It has 12 columns and 45 rows. The 4 top rows are for labeling and grouping columns, and the other 41 are dedicated 1 row for each treble note. The first column contains the names of the notes, the second and third columns are for entering the beginning bps for each bellows direction (as measured with reeds in the accordion), and the fourth and fifth are for listing the correction required (in bps) for each bellows direction in order to match the desired beat progression. The sixth column is just a reproduction of my cheat sheet, listing the cents per bps conversion factor for each note, and the seventh and eighth columns are for listing the calculated correction factor in cents for each bellows direction. This much I complete before taking the reeds out of the accordion.

        The last four columns are for listing:

    e) beginning cents deviation for each tongue as measured on the tuning bellows, and

    f) the tuning bellows target value for each tongue, which is derived by applying the cents correction factor to the beginning cents deviation for each tongue.

        To help keep the data properly organized and prevent errors, I use yellow highlighting to denote all columns pertaining to inner reed tongues. The completed chart constitutes a complete formulation of my musette reed tuning plan for one musette reed set.


Step 6: Tuning Musette Reeds

        Tune your musette reeds one musette reed set at a time. Before removing the reed blocks from the accordion, make sure each row of reeds is properly labeled, as discussed in “Wet or Dry?”, above. Establish the tuning bellows base line and the tuning bellows target value for each reed as you go, as detailed in e) and f) of step 5, above, then carefully clean each reed before you tune it. Check it on the test bellows again to see how much it improved due to the cleaning. Then tune each musette reed tongue to its tuning bellows target value, just as you did for the dry reed sets.

        Correcting errors and newly discovered problems – After all the musette reeds are tuned, put the reed blocks back in the accordion and check every musette reed pair again, in each bellows direction, making note of any needed corrections to the beat frequency on a fresh Musette Reed Tuning Worksheet. Use the cheat sheet column to convert each bps correction factor to cents. Establish a new tuning bellows base line for each reed tongue needing additional tuning, apply the cents correction factor to arrive at a new tuning bellows target value for each reed tongue needing additional tuning, and tune each of those reed tongues to its target value, just as you did the first time through.

        Occasionally, you may discover that the beat frequency for one of the pairs increases or decreases with changes in bellows pressure. Even though the higher note reeds are generally less prone to frequency drift in response to changes in bellows pressure, in the case of musette pairs the beat is so sensitive to minute changes in frequency that if one member of the pair drifts ever so slightly more than the other, the beat will change noticeably. If your final check reveals a reed pair whose beat varies noticeably with bellows pressure, then one of those reeds may have a loose rivet. I don’t mean “falling out loose”, or even “wandering off-center” loose. I mean just barely loose enough to allow a tiny bit more of the reed tongue to lift off the reed plate right next to the rivet whenever you get the reed tongue swinging wide by playing it at high volume. This increases the effective length of that reed tongue, lowering its pitch. In this case, you will have to identify which reed tongue is causing the problem and reset its rivet.

        If the musette set you are tuning is sharper than your clarinet set, a slower beat under high bellows pressure indicates the musette reed is going flat, while a faster beat under high pressure indicates that the clarinet reed is the culprit. If the musette set you are tuning is flatter than the clarinet set, then a slower beat implicates the clarinet reed, while a faster beat implicates the musette reed. In either case, if the problem occurs while opening the bellows, the inner tongue is causing the problem. If it occurs while the bellows is closing, the outer tongue is the one to repair.

        You can also identify which reed tongue you have to tighten by testing all four reed tongues individually on your test bellows with the Marcus Cavanaugh tuning app running on your iPad. Watch the readout as you vary tuning bellows pressure. If the frequency of any one of those tongues wanders more under bellows pressure changes than the other four, you will know to reset the rivet on that tongue.

        However you identify it, take the offending reed out of the reed block, place one end of the problem rivet on the edge of an anvil (just that rivet should contact the anvil, not the whole reed plate, not any reed tongue, and not both rivets) and give the other end of that rivet a couple of light raps with a hammer. Don’t hit it too hard, as you don’t want to distort the aluminum reed plate, which is relatively soft.

        Wax the reed back into the reed block, install it in the accordion, and test that pair again. If your hammer taps were sufficient, the problem will be solved, and the beat will be stable. However, it may not be stable at precisely the bps you want, in which case you will have to retune the reed tongue whose rivet you reset. With your cheat sheet in hand, it will be easy to arrive at a new correction factor in cents and a new tuning bellows target value, just as you did before.

        Musette Tuning tolerances - A look at the cheat sheet reveals that the tuning tolerances must be considerably tighter for musette reeds than for dry reeds, especially near the upper end of the 41 note scale. Tune to within two cents of your target at the low end of the keyboard, because at the lowest F, 2 cents equates to about 0.2 bps, which is as much accuracy as you need. Gradually decrease this margin of error as you ascend the scale, until at the very top tune to the nearest 0.2 cents, because at the highest A, 0.2 cents equates to about 0.2 bps. If each note is tuned to within 0.2 bps of your tuning plan it will sound perfectly tuned to anyone. If you let the tolerances wander to about 0.5 bps, some people (other accordion tuners, mostly) will notice. Most people are oblivious to beat frequencies, and will not notice anything wrong unless the beat is way off, in which case they will know it sounds bad, but they won’t know why.

Cleaning reeds

        The time honored way to clean reeds is by taking them out of the reed block, cleaning off the wax, and soaking them in benzine or carbon tetrachloride. That is not practical if you are not re-waxing the accordion, and in any case, I do not recommend using such toxic chemicals. Some people clean them in a liquid bath in an ultrasonic cleaner, but I can’t help wondering if there is any water in that liquid, and if so, what happens to the water that gets in between the reed plate and the rivet tongue all around the rivet. I prefer to clean mechanically.

        I use a tiny brass wire brush which I make by pulling individual tufts from a small wooden handled wire brush available at hardware stores (find a brush with the straight brass wire bristles, don’t buy one with curly bristles). Grip each tuft individually near the bottom with needle nose pliers and pull it out of the wood. Crimp a short length of steel tubing (I use 1/8” automotive brake line) onto each individual tuft to use as a handle, and snip the bristles to about 1/4" long, making a stiff narrow brush that fits down into the reed vent. Crimp one tuft in each end of the tube and on a few of them bend one set of bristles 90 degrees for working inside the reed block.



        These brushes work well for cleaning all but the underside of the outer reed tongue (unless you remove the reed from the reed block, in which case all surfaces are equally accessible). If you are working with a reed still mounted in its reed block, work from the outside, in order not to have to remove the interior leather. Lift the outer tongue and scrape the underside of it with a dull blade, such as the edge of a .004” or thicker metal feeler gauge, then pluck it to shake off the debris. This is not a perfect way to clean, but it is better than no cleaning at all. Then check its tuning again. After cleaning the reed in this manner you may have to file less than you originally thought, and sometimes you may not have to file at all. Sometimes cleaning alone brings the reed back into tune. Don’t clean it before you establish its tuning bellows target value, though, or you will have to put it back in the accordion to establish a new correction factor.

Aligning reed tongues

        If the reed is silent or flat because the reed tongue is out of perfect alignment with the vent, you will have to re-center it. With the reed removed from the reed block and held up to a bright background, verify visually that the spaces between the reed tongue and the vent are equal on both sides of the tongue. This takes a bit of skill, because if you do not look precisely straight on, or if your viewing angles from either side of straight on are not exactly symmetrical, one gap may seem wider when it is really not. Sometimes it helps to rock the reed tongue down into the vent before sighting through the gap. Claudio Binci routinely centers his newly made reed tongues by this method, and when he finishes, the clearance on either side of the tongue is exactly .0015”. Actually, the narrower the clearances, the easier it is to detect any asymmetry.

        The easiest way to re-center the tongue is to lever it back to the proper position. Insert a thin blade into the narrower space (midway along the tongue, so you can lift the tongue far enough to insert the blade, but not too close to the tip) and gently pry the reed tongue back to center. Take care not to twist the reed tongue or to damage the edge of the vent. If the reed tongue has worked its way off-center, it can only be because the rivet is too loose, so once you have it centered, you should reset the rivet, as described above, in “Correcting Errors and Newly Discovered Problems”. Naturally, you must wax the reed back into the reed block when you are finished.

Putting File to Metal

        Tuning is mostly about investigation, analysis, planning, organization, focus, and perspective, so it is appropriate that most of this tuning chapter has been dedicated to those critically important ingredients of tuning. However, that brings us to the remaining ingredients, which are skill, dexterity, and judgment, and this is where we finally put file to metal.

        Re-tuning in three stages - If you choose to re-tune, i.e., to alter the basic tuning by tuning to a different A, or to change a dry tuned accordion to wet (see the chapter on Tuning Musette Reeds), you should take the reeds out of the reed block for the first of three tuning stages. Re-tuning requires a lot of filing in the first stage, and adjacent reeds and leathers get in your way, making filing difficult. If you have a lot of filing to do, it is much easier to file a reed the same way the reed maker does, with the reed plate out of the reed block and lightly clamped in a vise, and with the leathers removed.


        The first tuning stage will be done with a file (except for the tiniest piccolo reeds), filing straight across the reed tongue (unless you are working on the tip, in which case you file directly toward or away from the rivet) to bring the reed to within a couple of cents of the final value. You need not obsess over small deviations at this stage, because the tuning will probably change anyway, perhaps by several cents, when you install the reed valves and wax the reed back into the reed block. You will not be able to compensate in advance for this change, because part of it is unpredictable. Sometimes the reed sharpens when mounted in the reed block and sometimes it flattens. Sometimes you get lucky and the change is very small or even zero.

        To check your progress during this first stage, place the reed directly over the aperture on the tuning table and draw air over the reed while managing your tuner. An assortment of different apertures will be required to fit the various reed sizes, so you must devise some way to fit those various apertures to your tuning bellows or tuning table. One way is to make a large opening in the top plate of your bellows and make an assortment of covers for this opening, each with a different sized aperture. One way to secure these covers is to slide them into slots cut around the top interior edge of three sides of a box mounted over the large aperture, as shown in the photo below. I’m sure there are other ways, as well, but in any case, you should line your apertures with leather, fuzzy side up, to maintain a good air seal between the reed plate and the bellows.


        The second tuning stage is also done outside the accordion, but with the reeds mounted (waxed) into the reed block and with all reed valves (leathers) in place. The goal at this stage is to bring the tuning to within a cent or two of the desired final value. You will use a file this time, too, but this time file diagonally so as not to damage the leathers on adjacent reeds (you will probably ruin a few leathers anyway, so have some replacements on hand). It would be pointless to attempt this tuning stage without the leathers in place, because installing them later would change the tuning again, requiring an additional round of tuning.

        The third and final tuning stage must be done in the accordion because installing the reed block into the accordion changes the tuning yet again. However, understand that tuning in the accordion does not necessarily mean actually filing or scratching the reed tongue while the reed block is mounted in the accordion, although this is sometimes done. For our purposes, tuning in the accordion means measuring the tuning deviation with the reed block mounted in the accordion, then taking the reed block out to make the necessary correction. This is because every reed tongue must be supported while you file it, and the reed tongues facing into the reed block can best be supported from the inside of the reed block, which is most easily and reliably done with the reed block out of the accordion.

        In this final stage you will be eliminating small deviations on the order of a few cents at most, so you may be using a scratcher, rather than a file (especially on the smaller reeds). If the needed correction is a cent or more on the largest reeds, or more than a few cents on the smaller reeds, you can start with the file (except on the tiniest reeds, where files are too crude), but once you have it within a cent on the largest reeds and within a few cents on the smaller reeds, switch to a scratcher to avoid overshooting your tuning target. On the tiniest piccolo reeds, both the file and scratcher are too course and crude, and you will be better off using only fairly fine sand paper, perhaps around #220. See the chapter on Tuning Tiny Piccolo Reeds.

        Spot Tuning - If you choose not to alter the basic tuning of your accordion, and if you are just tuning a few reeds that are noticeably out of tune with the rest, you will have a much easier time of it because in most cases you will begin with the third tuning stage. If a few reeds are out of tune by more than you can comfortably correct with scratching or diagonal filing, you can take them out of the reed block individually (after establishing the needed correction, of course), rather than having to re-wax the entire reed block.



        Tuning with a file - Tuning with a file is a fairly aggressive tactic, and requires a good sense of touch and pressure, a good memory for how many cents of change results from any given amount of file pressure on the various reed sizes, and considerable finesse. Files remove lots of metal, so typically not much filing is required to bring a reed back into tune. Make one light stroke and see what difference that makes to the tuning, and based on that, estimate how many identical light strokes will be required to bring it in line with your tuning bellows target value. Not all strokes are equal. With practice, you will learn to tell by feel how much metal you have removed with each stroke. If the file skips lightly over the reed, offering little resistance, you obviously remove much less metal than if it seems to dig in and offer significant resistance. One heavy stroke can remove as much as many light strokes. Until you develop a good feel for this, check your results frequently, perhaps after each stroke, in order to avoid overshooting your target.

        When filing an exterior reed tongue to flatten its pitch, file in the bottom third of the reed tongue. Insert a reed lifter or metal feeler gauge under the reed tongue to support it while you file. Extend the reed lifter or feeler gauge out over the adjacent reed valve to protect it from your file.



        When filing the tip of either an exterior or interior reed tongue, file directly toward or away from the rivet. Slip a flat reed lifter under the tip to protect the reed plate and adjacent leathers. Take great care to file only the top surface, not the end or sides, as the clearance between the reed tongue and the sides and end of the vent is critical, and you must not do anything to enlarge it. In handmade reeds and hand fitted reeds, this clearance is only .0015, so even a very light file stroke could double it, essentially ruining the reed. The smaller the reed, the more vulnerable it is to this kind of damage. On the very smallest piccolo reeds, where the reed tongue may only be .002” thick, you dare not use a file at all.


        In order to file the tip of an interior reed tongue, the tip of the tongue must be lifted up through the vent for access. This is done by inserting a small tool, wood, metal, or plastic, in through the opening in the bottom of the reed block and out through the tip of the vent. Then slip your flat reed lifter under the tip to support it evenly while you file. Be careful not to lever the reed plate out of the wax mounting during the process.



        This does not work on the smallest few clarinet reeds and on the top 20 or so piccolo reeds, because the tongue is too short to survive so much bending without suffering permanent damage. As a general rule, if the reed does not have leathers, it is probably too small to have its tongue lifted through the vent without permanently damaging it.

        In these cases, you can use a scratcher near the tip, taking special care to support the reed tongue evenly from below while you scratch. With this technique, you run the risk of cupping the end of the reed tongue due to excessive pressure and/or inadequate support from below. You also risk pushing burrs off the end of the reed tongue, where they will interfere with the tongue’s passage through the vent, either silencing the reed or throwing it out of tune. If this happens, use the edge of your reed lifter or of a .004” metal feeler gauge to scrape the tip clean. Alternatively, you can take the reed out of the reed block and tune it in your vise. The latter is by far the best course of action for the tiniest of reeds. 

        One of your objectives while filing must be to avoid creating stress concentrations that could lead to cracks. This means don’t cut a groove across the reed, not even a very shallow groove. Don’t remove metal from one tiny area, or from just one edge of the reed tongue. The wider the area and the thinner the layer removed, the better. Your file must be perfectly parallel with the top surface of the reed tongue, so it removes metal equally across the entire width of the reed. The best tuning would be invisible, blending in perfectly with the reed makers original file marks. This degree of perfection is practically impossible to achieve, especially when filing diagonally across the reed tongue, but it is nevertheless a worthwhile goal to keep in mind. The less conspicuous your file marks, the better.



        Tuning with a scratcher – When you are so close to your tuning bellows target value that you think one more file stroke might cause an overshoot, switch to using a scratcher. This probably won’t occur on the largest reeds, where it takes several light file strokes to make a cent of difference and where you are tuning to two or three cent accuracy, but on the smaller reeds, where you need more accuracy and where one file stroke can make several cents of difference, you will be using the scratcher often.

        Unless you choose to remove the reed from the reed block for tuning, you will also use a scratcher whenever flattening the tuning of the interior reed tongue, simply because you cannot get a file down into the vent. To flatten the pitch, use the scratcher to remove metal from the lower third of the reed tongue. If the cents change you need is more than you can get from a few scratches, remove the reed from the reed block so you can tune the originally filed side of it with a file.

        The number of cents change you can make with a scratcher varies with the size of the reed. On the largest reeds, you may only be able to get one cent or so without severely gouging the reed tongue, which is never a good idea. But on smaller reeds, you can use the scratcher to flatten the pitch by up to a several cents.

        To sharpen the pitch, you will use a file on the tip, after lifting the tip out of the vent for access. If the reed tongue is too small to lift out of the vent without permanently deforming it, then remove the reed from the reed block so you can scratch or file or sand the other side.

        Anytime you remove the reed from the reed block for tuning, establish a new tuning bellows base line for it by laying the reed directly on the opening in the top of the tuning bellows. Then apply your previously determined correction factor to get a new tuning bellows target value, which will be different from the tuning bellows target value you established for the reed mounted in the reed block.

        Tuning with a scratcher can be frustrating, because the changes in tuning you get are sometimes not those you expected, especially if your scratcher just moves metal around without actually removing it (see the discussion under “many pitfalls” in the Tuning Dry Reeds section).

        Scratching lightly sometimes does not remove metal, but simply lays over (flattens down) the ridges left by the reed-maker’s file. This is especially likely on handmade reeds, which are harder to scratch and which often exhibit more pronounced file marks. Laying over the file marks does not remove any weight, but it does shift the center of gravity in the direction the tool was traveling, and therefore impacts tuning, perhaps in the direction opposite to what the tuner intended if he was not aware of this subtle aspect of tuning.

        However, you can use this knowledge to your advantage in a way contrary to conventional wisdom by working any part of the reed tongue downward (i.e., toward the fixed end) to sharpen the pitch, and working any part of the reed tongue upward (i.e., toward the tip) to flatten the pitch. Thus you can, if you wish, sharpen the pitch by working with your scratcher on the lower third of the reed tongue, and flatten the pitch by working near the tip, exactly the opposite of the tuning procedure normally described. You just have to go lightly to be sure you are just laying the ridges over, rather than removing any metal. 

        This technique is especially useful where you need to change the tuning only slightly, say one cent or less, in which case removing any metal at all can cause you to overshoot your target, especially on the smaller reeds. It is also a much gentler way of tuning that avoids the higher pressures that can, if you are not careful, bend or arch the reed tongue in a way that adversely impacts the voicing or otherwise increases air leakage between the reed tongue and the vent.

        Naturally, this technique works best for the outer reed tongues, on which you have good access to the filed side. On interior reed tongues, unless you remove the reeds from the reed block, you normally work on the unfiled blue side of the tongue, where there are no previous file marks to lay over, so to use this subtle technique you would have to work through the small opening in the bottom of the reed block, where it is very difficult to see what you are doing. Due to those difficulties, I don’t normally try that. I scratch on the unfiled side unless I have taken the reed out of the reed block.

        Tuning with a Dremmel grinder - Some people tune with just a file and a scratcher, and maybe some fine sand paper, while some others prefer to use a Dremmel grinder. There is great controversy on this point, with some people even saying they would refuse to own an accordion that shows evidence of Dremmel tuning, because they believe the Dremmel grinder ruins the reeds.

        I have never seen any evidence that Dremmel grinding necessarily ruins reeds. The usual argument against it is that the Dremmel grinder removes too much metal. However, removing metal changes the tuning of the reed, so if too much metal had been removed, the reed would be out of tune. If the reed is in tune, that is strong evidence that exactly the right amount of metal has been removed. I have heard it claimed that if so much metal must be removed to bring it into tune, then the reed is no good anyway and should be replaced. However, I doubt that the person saying this has ever watched reed makers make reeds from scratch. They file very aggressively, removing copious amounts of metal. The amount of metal removed by a Dremmel grinder during a typical tuning is infinitesimally small in comparison.

        I have always tuned with just a file and a scratcher (and fine sandpaper for the tiniest reeds and for very small adjustments on some larger reeds) for the same reason that I always drill holes by hand. Power tools cut very quickly and it is easy to slip and drill or cut too far, and in an accordion, that can be especially damaging. Hand tools operate more slowly, and seem easier to control.

        However, I think Dremmel grinders offer certain advantages, and in the hands of a skilled operator, can probably be safe and effective. There is less pressure required, and thus less chance of bending the reed tongue or rotating it out of perfect alignment with the vent. The tiny grind stone is easier to maneuver in tight spaces than the relatively large and unwieldy tuning file, producing less chance of damaging the reed plate or adjacent leather. The high speed grinder removes metal evenly over a larger area, leaving a smoother, cleaner surface, free of scratches, gouges, file marks (stress concentrations that can lead to cracks), and free of the troubling burrs that can skew tuning results. In short, I think that while a Dremmel tool in the wrong hands can do a lot of damage, refusal to tolerate tuning with a Dremmel tool is probably rooted more in fear of change than in rational argument. I may experiment with a Dremmel on a junk accordion someday, to see if I can develop the skill, control, and confidence required to tune effectively with it.

Here is a reed that was apparently tuned (flattened) with a Dremmel tool:

This reed was tuned (flattened) with both a Dremmel tool and a file:

This reed was sharpened by filing at the tip, and also flattened by scratching near the base:

(Experienced eyes will detect that these are very old reeds, cut narrower at the base than modern reeds. In fact, these are from 1933.)

Tuning tiny piccolo reeds

        It is not uncommon to find the highest piccolo reed tongues (generally between the highest C and the highest A) arched, cupped, twisted, or kinked by excessive force during tuning or voicing. The problems are that these reed tongues are so small that it is very difficult to see what you are doing when tuning or voicing them, and they are so fragile that the usual tuning tools are inappropriate for tuning them. The smallest tongues are only about .002” thick near the free end. This is only 2/3 the thickness of a sheet of normal 20# copy paper, so they are very fragile. The teeth of a normal tuning file can catch an edge and tilt it upward, putting a twist in the reed tongue. The force necessary to make a scratch with a tuning scratcher can easily cup or arch the whole reed tongue if it is not very firmly and evenly supported from below. Lifting the reed tongue through the vent so it can be tuned from the outside would overstress it, permanently bending or arching the reed tongue. Attempts to lift or drop the tip to voice the reed can result in only the tip bending, rather than the whole tongue, as is required for proper voicing. Extreme care must be taken when tuning and voicing these reeds, or you will find yourself needing to buy new ones or to spend a lot of time straightening the old ones (see the chapter titled “Straightening Piccolo Reeds).


        I tune these tiniest reeds with small pieces of #220 sandpaper glued to the ends of small wooden rods. If the interior reed tongue needs tuning, I remove the reed from the reed block so I can tune the filed side of the tongue. I support the tongue during sanding by sliding a .002” metal feeler gauge under it, and for the very smallest ones, such as the one pictured, I use a .0015” feeler gauge. I establish the tuning bellows base line and measure my progress toward the tuning bellows target value by placing the reed directly on my tuning bellows, rather than putting it back in the reed block each time. Only after I have reached my tuning bellows target value do I wax the reed back into the reed block and put it back in the accordion for a final check.


Straightening tiny piccolo reeds

        If you need to straighten an arched, cupped, twisted, or kinked piccolo reed tongue, you will have to do it under fairly high magnification. The clearance under a properly voiced high A reed is about .0015” (yes, that is 15/10,000 of an inch). It is impossible to see that tiny clearance without appropriate magnification. It is likewise impossible to see arches, waves or twists of that magnitude, yet the reed must be perfectly straight and flat so the tiny critical clearance can be maintained along most of its length. If it is arched or twisted so part of the tongue is higher than that, excessive air will leak between the tongue and the vent, and the reed will respond slowly or not at all. If any part (other than right at the fixed end near the rivet) is lower than that, it will choke.

        I use a 15x binocular microscope and work under a bright light put very close to the reed. I clamp an anvil (really just a paper clip bent to the appropriate shape) to the base of the microscope, with the end of the anvil extending horizontally into the field of view, and at the proper height to be in focus (the depth of focus at these magnifications is very shallow). I hold the reed plate in one hand, a small bending tool (tiny screwdriver) in the other. I use the anvil to support the reed tongue near where I want it to bend, and apply force with the bending tool to make the bend. I normally rest the shank of the bending tool on the edge of the reed vent and rock the tool into the reed tongue in order to minimize the possibility that the tool might slip toward the free end and cause only the tip of the reed tongue to bend. If I am removing a kink, I slide a narrow strip cut from a .0015” metal feeler gauge under the reed tongue at the kink line to support it there while I press out the kink with either the bending tool or the anvil, or sometimes with both, one to hold the reed tongue firmly against the feeler gauge support, and the other to actually make the bend. To remove twists, I use two narrow strips cut from a .003” metal feeler gauge clamped together to allow slipping the reed tongue endwise between them, and I twist with that.

        These are extraordinary measures. If you have a new reed in stock or can easily get one, it is cheaper to install a new reed than to spend a half hour straightening the old one. But if getting a new reed takes time and if the delay is intolerable, then you might choose to salvage the old one. It can be done with the proper tools and patience. Naturally, after any significant bending and stretching, the reed is likely to be out of tune, so you should expect to have to tune it. However, a new reed will have to be tuned, also, so time-wise, that might be a wash.

Tools and supplies needed for tuning:

    Bellows pin pulling pliers

    Flat and phillips screwdrivers

    Tuning bellows

    Stroboscopic tuner, such as Peterson VS-II Virtual Strobe Tuner (about $300) or, if your accordion is tuned to A=440, and if you have an iPad or iPhone, you can use "Strobe Tuner Pro", an app for iPad and iPhone by Marcus Cavanaugh (about $3). Dirk's Tuner, a pc program available online as a download for $150, can listen to up to three reeds at a time and give you a report of the condition of each reed and the required corrections in either cents or hertz. (Caution: sometimes it hears only two of the three, and you can’t always tell which one is missing, so you don’t know which value is attributable to which reed.)

    Tuning file

    Tuning scratcher

    Reed lifter to support reed tongues and protect reed plates and leathers during tuning

    Reed wax

    Soldering iron modified for waxing (see article on waxing with a modified soldering iron)

    Reed leathers in all sizes

    Normally closed tweezers for handling leathers

    Sharp single edge razor blade for trimming leathers to proper length.

    Dull single edge razor blade for lifting and supporting the smallest reeds during tuning.

    Resin/alcohol glue (such as Indian Chief gasket shellac available at auto parts stores) for gluing leathers to reed plates.

    Naphtha or lighter fluid to remove spilled wax residue from reed plates and tongues.

    Soft cotton rag (with a small spot dampened with lighter fluid) for cleaning up spilled wax residue.

    Plastic, ironwood, aluminum, or brass scrapers for removing excess wax from reeds and reed blocks (avoid using steel scrapers on aluminum reed plates, as the relatively hard steel will scratch and damage the reed plates).

    Metal feeler gauges (for clearing stuck reeds, checking clearance between tongue and vent, and for supporting reeds while you file or scratch)


A final caution

        Tuning can be a mind-numbing repetitive process. The chart of tuning data for all the reed tongues can be a blur of columns and numbers. It is very easy to become bored, fatigued, distracted, absent minded, and careless while tuning, and this all too easily leads to mistakes. Here are some tips to help you stay organized and focused.

    1) Mark note values (note names) on the reeds or on the reed block or on a piece of tape attached to the reed block to help prevent tuning the wrong reed. Verify that your tuner indicates the correct name of the note you are tuning.

    2) Mark the name of the reed set (bassoon white, bassoon black, etc.) on each side of the reed block to help avoid inadvertently tuning a reed in the wrong set.

    3) Use worksheets or charts to keep tuning data organized. Make a separate worksheet for each reed set and label it clearly.

    4) Make worksheets illustrate the process and numbers used to arrive at the tuning target for each reed in order to facilitate double checking all calculations.

    5) Highlight or color code all chart data relating to inner reed tongues to help distinguish them from data relating to outer reed tongues.

    6) Use plus signs on all correction factors for reed tongues that need to be sharpened and use minus signs on all correction factors for reed tongues that need to be flattened.

    7) Circle each reed tongue's tuning target value as you finish tuning that tongue, in order to reduce the likelihood of inadvertently using that target value again for some other reed.

    8) Worksheets should show the initial tuning bellows reading for each reed tongue alongside the tuning bellows target value to facilitate keeping track of progress toward the target value in order to better judge how much additional filing or scratching will be required. This will help avoid overshoots.

    9) Be cautious and alert for errors. Double check all calculations. If a target tuning value does not seem right, that is, if it is very different from the target values for neighboring reeds, then double check everything before putting file to metal. It is far easier, and far less disheartening, to correct mistakes on paper than on the reeds themselves.

The full explanation of the formula for determining cents per bps for each note.

        An understanding of exactly how cents relate to hertz (hz), why each cent equates to more hertz at the higher notes, and how to calculate the exact number of hertz that equates to each cent for any note, begins with an understanding of our 12 note equal temperament chromatic scale and the fact that with each octave, the frequency in terms of hertz doubles.

        At piano A1 the frequency is 55 hz. An octave higher at A2, the frequency is 110 hz, at A3 it is 220 hz, at A4 it is 440 hz, at A5 it is 880 hz, at A6 it is 1760 hz, and at A7 it is 3520 hz. Clearly, each of the 12 equal semitone intervals in any higher octave is spread over a greater number of hertz than its counterpart in any lower octave. Since each semitone interval occupies a greater range of hertz in the higher octave, it follows that each cent (1/100 of a semitone) also occupies a greater range of hertz in the higher octave.

        We label basic tunings in terms of the note A (A=440, etc.) because only the A notes in our scales have integers (whole numbers) as their frequencies. All the other 11 note frequencies in each octave are then derived from the frequency of the A according to a specific mathematical formula that divides each octave into 12 equal pitch intervals.

        This formula is exponential, rather than linear, because our perception of pitch is not linearly related to frequency. The relationship between frequency and our sense of pitch is exponential (a linear relationship would imply that each successive octave interval would occupy the same number of hertz, rather than doubling each time). This holds true within each octave, as well. That means we cannot simply divide each octave into 12 equal numbers of hertz, but must instead assign each successively higher note within the octave a greater range of hertz than its predecessor, or the intervals would not sound equal.

        The formula that does this perfectly is an exponential formula based on the fact that 2 raised to the 1/12 power multiplied by itself 12 times = 2. That is,

21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12 * 21/12
= 2(1/12+1/12+1/12+1/12+1/12+1/12+1/12+1/12+1/12+1/12+1/12+1/12) = 21 = 2

        By assigning each successive note on the chromatic scale a frequency equal to the frequency of the preceding note x 2 raised to the 1/12 power, after 12 successive notes we arrive at a frequency equal to 2 times the original note. In other words, this formula doubles the frequency at each octave, while making each semitone pitch interval sound equally spaced. Here is the formula applied to calculate the frequencies of each note in the octave immediately above A4:


A# = 440 x 21/12 = 440 * 1.059463094 = 466.1637614

B = A# x 21/12 = A * 22/12 = 440 * 1.122462048 = 493.8833013

C = B x 21/12 = A * 23/12 = 440 * 1.189207115 = 523.2511306

C# = C * 21/12 = A * 24/12 = A * 21/3 = 440 * 1.25992105 = 554.3652619

D =

D# =

E =

F =

F# =

G =

G# = A * 2 11/12 = 440 * 1.887748624 = 830.6093947

A = G# * 2 1/12 = A * 2 12/12 = A * 2 1 = 440 * 2= 880 Hertz

        Just as each successive semitone interval within an octave spans a greater range of hertz as we ascend the scale, so does each cent interval within any semitone span a greater range of hertz as we ascend through 100 cents to get to the next note in the scale. Thus, the difference in terms of hertz between A4 and a tone one cent sharper than A4 is less than the difference in hertz between the two tones 25 cents and 26 cents sharper than A4.

        The formula that expresses this relationship perfectly is very similar to the one used for assigning frequencies to notes within an octave, and is based on the fact that 2 raised to the 1/1200 power multiplied by itself 100 times = 2 raised to the 1/12 power. In other words,
100 cents = 2100/1200 = 2 1/12 = 1.059463094. So +100 cents from note A becomes A# = 440 x 2 1/12 = 440 * 1.059463094 = 466.163761 hz. Thus, advancing 1 cent at a time for 100 cents (i.e., a semitone) brings us exactly to the next note in the scale.

        This formula provides that a tone x cents sharper than any given tone has a frequency equal to that of the given tone multiplied by 2 raised to the x/1200 power. For example, the frequency of a tone 1 cent sharper than A4 is 440 x 2 raised to the 1/1200 power. Similarly, the frequency of a tone 5 cents sharper than A4 has the frequency 440 x 2 raised to the 5/1200 power. Thus, the frequency of a tone 100 cents higher than A4 is 440 x 2 raised to the 100/1200 power, which is the same as 440 x 2 raised to the 1/12 power, which is the formula for the frequency of the next note up the scale, A#.

        Here is the formula applied to calculate frequencies for some tones various numbers of cents sharper than A4:

1 cent = 21/1200 = 1.000577790. So +1 cents from note A = 440.2542276 hertz

5 cents = 25/1200 = 1.002892288. So +5 cents from note A = 441.2726067

10 cents = 210/1200 = 1.005792941. So +10 cents from note A = 442.548894

15 cents = 215/1200 = 1.008701984. So +15 cents from note A = 443.8288728


    To calculate frequencies of tones various numbers of cents flatter than A4, we take the reciprocal. In other words, 

-1 cent = 1/cent = 1/21/1200 =1/1.000577790 = 0.999422433, so -1 cent from note A = 439.000577567  

-5 cents = 1/25/1200 = 1/1.002892288 = 0.997116053. So -5 cents from note A = 438.7310634

-10 cents = 1/210/1200 = 1/1.005792941. So -10 cents from note A = 437.4657865

-15 cents = 1/215/1200 = 1/1.008701984. So -15 cents from note A = 436.2041584


        Notice that the number of hertz equating to -5 cents (-1.2883 hz) is not exactly equal to the number of hertz equating to +5 cents (+1.2726 hz), but they are so close (within .0157 hz) that no one would ever hear the difference, not even on musette tuning at the highest notes. That is why it is okay to use the simplified "cheat sheet", which uses a single conversion factor for tuning either sharp or flat. No one can hear a .0157 bps difference in beat frequency.

        Now you know how to calculate the difference in hertz between two notes any given number of cents apart. Since each hertz of difference generates 1 beat per second (bps) when the two notes are played together, you now also know the number of bps that is generated for any given number of cents difference. A very clear, plain language explanation of why each hertz of difference generates one bps can be found in Chapter 27 of PIANO ACCORDION OWNER’S MANUAL AND BUYER’S GUIDE.


To find out more about Piano Accordion Owner's Manual and Buyer's Guide, click here.

To order it, click here

To return to the top of this page, click here. 

$25 plus shipping and handling:
$5 to U.S. addresses
$16 to Canada
$23 to all other countries

Find out more about it here.

Order it here.

Care to share Accordion Revival with your friends?

Website Builder