The Origins of Tone: An Abridged History of the Piano Sound
I remember the specific day that it struck home to me how personal and personalized musical instruments have always been. The year was 2016, and I was visiting Germany on an internship through my university. Surrounded by about 70 different kinds of horn instruments, I pondered how ordinary people who also tilled fields, hunted for sustenance, or engaged in early trades, also built musical instruments that fit their lifestyles and their music. A carpenter might make early harps or violins, a farmer carve reed pipes or hollow out rams horns, a blacksmith might also make bells or drums. In archeological sites across the whole world we find musical instruments made of the most readily available materials, whether of reed or bone, brass or wood(1). The whole of human history has truly been full of music.
The piano is a relatively modern invention (around the year 1700) and as an instrument too large to be strictly portable, and too demanding in its construction to be easily accessible by those of ordinary means, was more a toy of the elite than a means of expression for the masses. However, even at that early point in time, keyed instruments were already becoming more widely accessible in the form of the clavichord (2).
Beginning in the early 1400s, the clavichord and the (perhaps more well-known) harpsichord, filling very distinct musical roles, contributed uniquely to music in ways that paved the way for an instrument like the piano to come into being. The harpsichord could be heard through an orchestra, while the clavichord was used in small-group and home settings. Though never as popular as its larger cousin, for centuries the clavichord saw use by composers such as CPE Bach and Mozart either as performance or as practice instruments. (3)
Bartolomeo Cristofori designed the piano to be a sort of harpsichord with changes to the action that let it play in a much wider dynamic range, from extremely quiet to loud and dramatic. This is quite evident in the tone of the earliest pianos by Cristofori and later Silbermann, who soon after innovated on the original design. These proto-pianos are much closer in tone to the harpsichords from whence they were developed than they are to modern pianos, though in the two octaves above middle C, that clear piano-like sound was already beginning to develop.
In order to understand the tone of a piano, we may start by examining the way strings produce sound and how steps have been taken to beautify that sound.
In his 1877 work “On the Sensations of Sound” Hermann Helmholz discussed at length this exact history of the usage of strings in the creation of music, and the various tone qualities exhibited by them. For modern pianos, each string when played, is incited to play various wavelengths at the same time, the deepest of which is perceived* to be the actual note played, with higher wavelengths being processed by the human ear as “color”. Just as it is often difficult even for someone familiar with all of the sounds of each orchestral instrument to pick them out from each other in an ensemble in real time, our ear tends to hear each portion of any sound containing multiple overtones only as “color” on the whole.
Helmholtz describes an experiment in which he prepared two glass bottles and air hoses to blow over each, one playing a B flat, and the other a B flat one octave higher. When played alone, the first bottle made that familiar “oo” sound that the opening of a bottle neck can produce. However, when the second bottle sound was added, the ear was able to discern them as separate tones only briefly and with effort before they seemed to flow together into one note. Helmholtz described the resulting tone to have the same pitch as the lower bottle, but with a changed tonal character, seeming to make an “o” sound instead of “oo”, such as in the word “spoke,” rather than “spook.” Helmholz further describes that the full overtone series that appears in a struck string may not be perceived as individual tones themselves, but rather change how we perceive the overall tone quality of the string. (4)
One might be tempted to think that the more partials that a string can create, the better the tone quality will be and the more pleasant the overall sound of the string, however, experience has shown this not to be the case. In fact, one major problem with the way harpsichord strings were plucked was that higher overtones overpowered the lower tones, so much so that the sound would become harsh. In fact, the ability of metal to transmit very small, very high waves is what gives sounds produced by metals a typical “metallic” twanging sound. Although improvements with the purpose of mitigating these overtones were tried on harpsichords into the early 1800s, by that time the harpsichord was already in the process of being replaced on the concert stage, and development in the area slowed. (5)
By contrast, during the later 1700s, development in piano design was making great strides. By incorporating aspects of the dulcimer into harpsichord hammers, a stringed instrument of the time where a player used mallets to strike strings the way a modern xylophone player might, the harshness in tone relative to plucked harpsichord strings was much reduced. Piano manufacturers experimented with various materials, such as hollow circles made from paper and glue that featured on the original Cristofori designs, to leather or buckskin on pianos designed later in the century, up until the eventual adoption of modern felt hammers in 1826. One commonality shared by all of these materials is a certain amount of shock absorption, or springiness. Essentially, a softer hammer spreads the shock of impact into the string over a greater area of the string, which inhibits the production of the smallest and therefore highest overtones. Controlling the softness of piano hammers became a great tool that allowed manufacturers to shape the sound of the piano more precisely, and after years of further refinement, the modern piano sound was achieved.
Or at least, part of it was. Of course, hammers are not the only element of the sound of a piano. One additional aspect of piano tone we will discuss here: the strings themselves are another major source of good piano tone. In early pianos, the bass strings in particular sounded very different from how a modern listener would expect. I recommend this demonstration video from the YouTube channel Baroqueband as an example of what this sounded like.
https://m.youtube.com/watch?v=a9IaE2i-DmA
As you may hear, the notes on the treble side of the piano sound much closer to a modern instrument, where the bass sounds more nasal and reminiscent of a harpsichord. On the modern instrument, the bass notes sound much more resonant, and part of this is due to their length, thickness, and the tension to which they are subjected. Already these pianos were utilizing higher string tension than the harpsichords, and even iron or in some cases steel in the highest notes, something harpsichord makers usually eschewed, favoring instead the softer and more fragile brass at lower relative tension. It is these brass strings at low tension in the bass that give the example piano its tinny, harpsichord-like quality.
What determines if a piano sounds “tinny” or “resonant,” having “good tone” or “shrill and unpleasant” is due several factors including hammer hardness, the position at which the hammer strikes the string (certain parts of the string will favor certain resonant overtones, and standardizing strike point to between 1/7 and 1/9 of string length helped builders accomplish this in a significant way. More on that in a future article.) As any piano technician with restringing experience knows, another important piece of the piano tone puzzle comes from the strings themselves.
In the early 1800s, cast iron frames and improvements in steel manufacturing allowed for high quality steel strings to be used at high tension. This was desirable because of the way tension interacts with string physics; metal strings become more elastic at higher tension and louder, giving manufacturers more tools in balancing the overall volume of the piano and the fine-tuning balance between its various sections (ensemble musicians know how important balancing loudness is between different parts of a band or orchestra). Crucially, the most important aspect of piano physics: the attack and initial decay of sound, is heavily tied to the physics of the strings themselves. (6) When strings are inelastic, either by design or through crystallization of the metal due to age, each wave vibration becomes inhibited by a (usually very small, but increasing for each higher overtone) percentage, that causes the very small topmost overtones to sharpen relative to the fundamental pitch, becoming “noisy” or “discordant”.
New steel strings in modern pianos tend to have advantages in tone, color, and overall loudness relative to older strings and historic scales.
One final consideration for this article, the majority of the sound information is contained in the initial strike, followed by a more slowly decaying long wave sound. The waveform of the piano strike changes rapidly over the first few fractions of a second, as the string creates a range of frequencies on being struck before it stabilizes into a regularly repeating, if slowly diminishing, sound wave.
It is for this reason that realistic piano sounds have proven difficult to reproduce mathematically, a problem that plagued the early years of sound synthesis. Using mathematics to model a struck string is itself straightforward, as was described by Julius Cook in his landmark studies in sound synthesis at Stanford in the 1980s(7), however the sound of a single note struck on piano contains much more nuance, including elements of noise from the key hitting felt, the dampers raising and falling, and alterations in the wave from the string stabilizing, variations due to slight detuning of unison strings, and also sympathetic resonance from other strings in the treble portion of the piano.
Early explorations into a subtractively synthesized piano sound struggled with the difficulty that early synthesizers had with memory. These instruments weren’t able to store enough information to reliably capture the complex attack and decay profile of the piano sound. At that time, the scientific foundations of piano were still in the process of being explored. The trial-and-error based approach that led to the formation of the modern piano tone had only sporadically been the subject of serious scientific inquiry and was not always understood, even by those who were building the instruments.
The first relatively realistic sounding piano synths were created by digitally processing the recorded sound of acoustic piano sounds, since real-time processing of realistic piano sounds would have been too processing intensive at the time and were not adequately understood even at a base level.
Similarly, the best patch-based piano sounds are painstakingly sampled by high-quality microphones, listening to real world pianos. There hasn’t been at the time of writing a publicly available synth based piano sound that has been able to match sample-based piano sounds in terms of tonal depth and realistic response, though some have tried. (8) For the most part, musicians are more interested in if your synth can make something that sounds good than simply being technically unique, a lesson Yamaha learned the hard way with their digital pianos back in the 90s,(9) and development in that area also seems to have diminished.
(I was very proud of this Yamaha PSR 195 Digital Piano back in 2003. Technology has come a long way since then)
That being said, although the concert piano in its current form has remained outwardly similar in the last 100 years, home pianos have undergone huge shifts in response to changing market conditions and trends. Many great piano makers have come and gone in the Midwest and left their stamp on the industry, building instruments that have taken, used, stretched, innovated, and of course, occasionally ignored the fundamental pillars of good piano tone. Some significant innovations to piano design in the last century include the use of modern material science to replace the ivory and enameled ebony in keytops.
Modern keytops are cheaper, harder, do not discolor or chip in the same way typical of ivory, wood or bone, and last longer. Teflon has been used to aid movement at the turning points of some modern piano actions, which unlike wood, are resistant to slowness caused by swelling. In terms of future innovations, some way of producing rustless strings would be useful. Up until this point the elastic limit of stainless steel has been too low to accommodate the tensions in modern pianos, but further advances in material sciences could also facilitate improvements in piano design in the future. (10)
Returning to the museum of musical instruments, the state of Missouri in the years I have been able to travel through seems itself to be a grand museum of piano history, with instruments of all kinds and variations, in homes, churches, schools, venues, and more. I am honored to be able to visit them and visit you, in order to appreciate so many of them in homes all across the state. I look forward to meeting you and helping your piano play to its full potential!
*in the deepest bass notes on a piano, the fundamental tone may be outside the range of normal human hearing, yet still is perceived through the implication of its higher overtones.
(1)Encyclopedia Britannica entry under musical instruments
(2)Men Women, and Pianos: A Social History. P. 221
(3)https://www.nytimes.com/2006/08/30/arts/music/30moza.html “Knowing Mozart Better in the Evolution of the Piano”
(4)On the Sensations of Tone, Hermann Helmolz, Dover Books, 2016, p. 60
(5)”Pianos and their Makers” Alfred Dolge p.37
(6) “The Piano, its History, Makers, Players, and Music” p. 35
(7) https://ccrma.stanford.edu/~jos/pmupd/Strings.html
“Physical Modeling Synthesis Update”
(8)https://www.youtube.com/watch?v=uuebeNV-DS8 “Synthesizing an Acoustic Piano - progress” on YouTube by 9b0
“Chapter 4:Changing Needs & A Return To Roots”
(10) “The Piano - Its Acoustics” W.V. McFerrin, p. 14-18
