Part 1: Measuring the "Unmeasurable"
By Jason Cassel, RPT
The Steinway World-Wide Technical Reference Guide recognizes voicing as the “greatest opportunity for the technician to become creative,” adding that “every instrument should be balanced and melodic, and have a variety of color.” With this declaration, however, comes the conclusion that “voicing is not measurable.” This is perhaps due to the fact that tone-building is comprised of a variety of factors, including the piano’s construction, the material and shape of the piano hammers, the regulation of the piano action, and even the acoustics of the room.
In addition to these challenges, André Oorebeek, in his book The Voice of the Piano: A Piano Technician’s Definitive Guide to Voicing, adds, “Verbalizing [the voicing] process is perhaps even more difficult because talking about sound is an abstraction everyone may interpret in a different way.” While I agree that voicing is difficult to measure and define, I believe that there is a great deal that piano technicians can learn by attempting to analyze the subtle changes in the piano’s tonal characteristics produced by voicing techniques.
I have chosen to focus my study on what many consider to be the “final steps” of tone-building, namely, hammer needling and the application of hardening chemicals to the hammer. While recognizing that this emphasis is limited (no amount of hammer needling can compensate for a poorly regulated action), these techniques are generally what come to mind when the term “voicing” is mentioned. Stephen Brady, in his book Under the Lid: The Art & Craft of the Concert Piano Technician, affirms, “As long as the instrument’s underlying structure is sound and the preliminary work has been done properly, hammers are indeed the place where we can make the most dramatic changes in a piano’s personality.” Brady adds his own definition of voicing as “the adjustment of a piano’s tonal characteristics by manipulating the shape, mass, density, and stiffness of the hammer felts.”
To perform this study, I began by sampling worn hard-pressed and soft-pressed hammers of different manufacturers and reshaping them. I then took these sample hammers into a recital hall at Florida State University, where I swapped them with hammers in two concert pianos. To examine these hammers where they would most likely be used, I placed the hard-pressed hammers in the Asian-made piano, and the soft-pressed hammers in the American-made piano. I then made adjustments to action regulation as needed.
Microphones were set up and an audio engineer recorded these sample notes on each piano before and after voicing techniques were performed. To ensure that the same amount of force was applied each time, a colleague of mine assisted in creating a device similar to the “thumper” used in the RPT tuning exam, where a weight is dropped from the same height each time.
Once the notes were recorded, I was able to analyze the partials through the use of a spectrum analyzer. At this point, I needed to address yet another challenge in how to best measure the effects of voicing, which is that the individual volumes of partials change with time. From February to June of 2018, David M. Koenig published a five-part series in the Piano Technicians Journal entitled “Spectral Methods for Piano Analysis.” These articles featured stunning graphs that represented changes in partial amplitudes and other aspects of tone over time. In the study reported here, however, I have chosen to analyze only the attack of each note for the sake of simplicity. This was done by taking a snapshot of the note’s spectrum a few milliseconds after the moment of the attack. Once again, I recognize that this approach is limited in its scope. That being said, I believe that doing so makes the graphics easier to understand and therefore more useful to piano technicians.
The effects of voicing are often defined by generalized statements, for example: “Doing this brings out the higher partials,” or “This technique adds body to the tone.” I would like to attempt to define these changes more clearly. Before doing so, however, Iwillfirstdescribehow I will refer to the different regions of the hammer. Each technician has his or her own way of defining these areas. Again, for simplicity, I will define these regions as, Low Shoulder, High Shoulder and Crown, with the exact center of the crown being referred to as the Strike Point.
I will define gradations of needling by length. I will refer to Shallow Needling as the use of needles 3-4 mm long, and Deep Needling as the use of needles 5-10 mm long. And finally, I would like to attempt to define tone itself.
Timbre or tone color is defined in the Webster dictionary as “the quality given to a sound by its overtones.” In other words, the timbre of a musical instrument is defined by the volume of its overtones. This is why a trumpet playing A440 sounds different than a clarinet playing the same pitch, or why a single note on a xylophone played with a yarn mallet sounds different than it does when played with an acrylic mallet. Piano technicians refer to these overtones as partials.
Because a piano hammer is manufactured with the outer surface softer than the inner layers, the hammer functions as a type of “nonlinear spring.” Mario Igrec, in his book Pianos Inside Out: A Comprehensive Guide to Piano Tuning, Repairing, and Rebuilding, describes this process, explaining, “A felt hammer acts as a relatively soft substance during soft blows (similar to a tennis ball), but as a much stiffer substance during hard blows (more like a baseball).” Stephen Brady adds, “This nonlinearity contributes to the change in tone color between soft and loud playing.” Finally, Mario Igrec defines this tonal change more specifically, stating, “The hammer head is a medium of variable stiffness: in soft playing it damps high partials, whereas on hard blows it releases their full spectrum. This creates a tonal gradient that makes the piano as expressive and versatile as it is.”
To examine this timbral change over the dynamic range, I played the same note from pianissimo to fortissimo, increasing the dynamic level with each blow. Now, admittedly, this process is subject to a fair amount of human error. All the same, the graphs produced exemplify this tonal gradient quite well. Graph 1 represents the tonal gradient of a hard-pressed hammer playing F4 at five dynamic levels (from pianissimo to forte). Graph 2 represents the same tonal gradient with a soft-pressed hammer.
Graph 1: Tonal gradient of a hard-pressed hammer playing F4 at five dynamic levels.
Graph 2: Tonal gradient of a soft-pressed hammer playing F4 at five dynamic levels.
This video presents the hard-pressed hammer graphic with its respective recording:
In these graphs, the x-axis shows the partial number and the y-axis shows the volume of each partial. The color of the line represents what dynamic level was played. For example, notice how the light blue line in both Graph 1 and Graph 2 starts high and then drops steeply until bottoming out. This means that when that note is played at pianissimo (light blue line), at the moment of attack, the lower partials have some volume and the higher partials do not. Contrast this with the dark blue line. This line represents the note played at forte. Notice how the dark blue line does not drop down or bottom out. This means that when this note is played at forte, the lower and higher partials have volume.
With this in mind, comparing Graph 1 and Graph 2 reveals some surprising similarities. In both, the pianissimo and piano dynamic levels engage most prominently the first 8-9 partials (the light blue and orange lines). In the mezzo-piano and mezzo-forte range, the first 20 partials are engaged – with some higher partials already being engaged in the hard-pressed hammer (the grey and yellow lines). At the forte dynamic range, the full tonal spectrum is engaged – again, with the hard-pressed hammer demonstrating higher partials comparatively louder (the dark blue line). This more notable introduction of higher partials is characteristic of hard-pressed hammers. André Oorebeek explains in general terms, “A very hard hammer causes mainly a higher partial sequence and a soft hammer mainly a lower partial sequence.”
As mentioned, voicing is often defined by generalized statements. Even the quotations used in this section generalize color changes in the undefined terms of “high partials” or “low partials.” From Graphs 1 and 2, I believe that these ranges are best defined as follows:
1. Low Partials: Approximately partials 1-9 – Defining the core or power of the tone.
2. Middle Partials: Approximately partials 10-20 – Defining the body or presence of the tone.
3. High Partials: Approximately partials 21+ – Defining the color or attack of the tone, as determined by their absence or presence.
These approximate ranges only apply to the midrange and bass range of the piano. In the treble ranges, the core and power of the tone is generally built in the first 1-3 partials, the body and presence of the tone is found in approximately partials 4-7, and any partials higher than 7 would constitute the color and/or attack of the tone through their presence or absence. Obviously, in the extreme treble, these numbers would likely be condensed further.
Graph 3: Low, middle, and high partial ranges at F4, hard-pressed hammer.
In their eBook Complete Piano Voicing, Jim Busby and Vincent Mrykalo explain, “At different dynamic levels certain ‘tone colors’ should be heard, and the voicer must be aware of the piano’s tone at every volume.” With this in mind, in Video 2, I used an equalizer to isolate each of the partial ranges defined above (low, middle, and high as shown in Graph 3). Listening to these ranges independently is extremely helpful in distinguishing when and where changes are made when voicing. For example, after watching Video 2, try watching Video 1 again to see if your ears can recognize when each partial range is added.
Before moving forward, allow me to add that my purpose is to not to draw any definitive conclusions, but rather to observe. Clearly, my collected data from a handful of hammers on two pianos is statistically insignificant. I will therefore not speak in terms of better or worse, or right or wrong. I will simply make observations drawn from my findings and the research of others.
Mating Hammers to Strings
All piano and hammer manufacturers agree that mating hammers to the strings is an essential step in tone-building and regulation. This means that the hammer hits all three strings (or two in the bichord section) at the same time. Stephen Brady explains, “To deliver maximum power and clarity of tone, the hammers must contact all strings of their respective notes together, so the resulting vibrations of the three strings will be in phase with each other.” When strings vibrate out-of-phase from one another, certain partials may “cancel out.”
To mate a hammer to the strings on a grand piano, the hammer is lifted up until it blocks against the strings. This is generally accomplished by using your finger to push up on the jack tender. The strings are then plucked with a fingertip or guitar pick while the hammer is lightly touching the strings. If a string rings out, then the hammer is filed accordingly until all strings are muted when plucked. Note: The strings must be level first.
Graph 4 shows the tonal spectrum for F6 played with a hard-pressed hammer before and after the hammer was mated to the strings.
Graph 4: Partial series on F6 before and after mating hammer to strings.
As you can see, the amplitudes of partials 4-6 (which constitute much of the body and presence of the tone in this region of the piano) were significantly increased. Additionally, all other partials were also increased. This is due to the strings now vibrating in phase with one another. Video 3 includes a side-by-side comparison of the recorded audio analyzed in this section:
In the first capo section of some grand pianos, notes tend to die off rather quickly when sustained. To determine the potential of the belly of the piano, a string is generally plucked hard with a fingertip or guitar pick and the sustain time is measured. The same note is then played, and the resulting sustain time is compared to the potential. If the plucked sustained time is greater than the played sustain time, then this signifies to the technician that improvement can be made. These improvements are often accomplished through bedding the keyframe and regulating the action. However, if these steps still leave room for refinement, then the Renner technical manual (Selecting & Voicing the Renner Hammer) suggests shallow needling up the bass and treble sides of the hammer with a chopstick voicing tool. This technique can be performed with the action in the piano by putting the hammer in check before needling.
Suggested shallow needling points.
Graph 5 shows the before and after effects of this technique applied to a hard-pressed hammer playing F6. The box on the right side of the graph shows the sustain times. Notice how the sustain time was increased without drastically altering the tonal spectrum for the note. As a point of interest, the soft-pressed hammer showed a slight increase in sustain time, but overall, it remained unaffected.
Graph 5: Effect of shallow needling up bass and trebles side of hard-pressed hammer at F6.
Conclusion of Part 1
I hope this introduction has shown that while voicing may be difficult to measure accurately, there is much to be gained by taking a more analytical approach to a subject that is often veiled in obscurity and generalizations. In Part 2 of this series, we will examine voicing techniques commonly used to build power and tone, using graphics and recording to measure findings.
Baldassin, Rick. Selecting & Voicing the Renner Hammer. Scottsdale, AZ: Renner USA, 2016.
Brady, Stephen H. Under the Lid: Tthe Art & Craft of the Concert Piano Technician. Seattle: Byzantium Books, 2011.
Busby, Jim and Vincent Mrykalo. Complete Piano Voicing: Pragmatic, Intuitive and Inclusive. Piano Technician Tutorials, 2018.
Igrec, Mario. Pianos Inside Out: A Comprehensive Guide to Piano Tuning, Repairing, and Rebuilding. Mandeville, LA: In Tune Press, 2013.
Koenig, David M. “Spectral Methods for Piano Analysis.” Piano Technicians Journal vol. 61, issue 2 (February 2018): 20 – 23.
Merriam-Webster. “Timbre.” Merriam-Webster. https://www.merriam-webster.com/dictionary/timbre (accessed 4/24/18).
Oorebeek, André. The Voice of the Piano: A Piano Technicians Definitive Guide to Voicing. Nanaimo, BC: Crescendo Publications, 2009.
Russell, David A. “The Piano Hammer as a Nonlinear Spring.” GMI Engineering & Management Institute. http://www.acs.psu.edu/drussell/piano/nonlinearhammer.html (accessed 4/24/18).
Steinway & Sons. World-Wide Technical Reference Guide. New York: Steinway & Sons, 2017.
Jason Cassel, RPT, M.A.
Brigham Young University - Salt Lake City, UT
Jason Cassel currently serves as a Piano Technician for the Brigham Young University School of Music. Before accepting this position, Jason Cassel graduated with an MA in Piano Technology from Florida State University – a distinction shared by only a dozen technicians in the nation. He earned his BM in Commercial Music from BYU while working as a student apprentice in the university’s piano shop. Jason has received training at the Steinway & Sons Factory in New York, the Mason & Hamlin Factory in Boston, and from Yamaha Pianos in California.
Jason is a Registered Piano Technician with the Piano Technicians Guild and has presented at local chapter meetings and attended various regional and national conventions. Many of his articles on piano technology have been published in the Piano Technicians Journal and his work on the revolutionary On Pitch DVD Series has been considered “an unprecedented tool to propel the next generation of fine tuners.” (Anne Garee – Former Head of Piano Technology at FSU)