In the Wind . . .

December 23, 2020

How does it work?

It happened again. I sat at this desk for days mud wrestling with an unruly topic for this column. Twice I had more than a thousand tortured words on the screen, went upstairs for a break, and came back to Ctrl-Shift-A-Delete. But Anthony Tommasini, music critic for The New York Times, came to my rescue with his article under the headline, “Why Do Pianists Know So Little About Pianos?,” published November 12, 2020. This article was born as the outbreak of COVID-19 got rolling in New York City last March and his piano needed tuning, but his apartment building was locked down and workers from outside were not allowed in except for emergencies. “An out-of-tune piano hardly seemed an emergency.”

He quotes the brilliant Jeremy Denk as not knowing “the first thing about piano technology.” Denk, whose playing I admire deeply and who like me is an alumnus of Oberlin College, had the same issue as Tommasini when his building locked down, but convinced the superintendent of his apartment building that because playing the piano is his profession, his tuner should be accepted as an essential worker. It worked.

Tommasini singles out Mitsuko Uchida as one prominent pianist who is an intimate student of piano technology. He quotes her as saying, “you get stuck when the weight is different key to key, the piano has been sloppily prepared, and the dampers have not been adjusted—or the spring in the pedal.” She went on, finding trouble when “the pin underneath the key [guide pin] is dirty, or the other pin in the middle of the mechanism [balance pin] is dirty, rubbing, or slurping.” I love the word slurping in this context.

Tommasini reminds us that orchestral players know more about their instruments than most pianists, and that unlike pianists, orchestral players own their instruments and can carry them with them between performances. Vladimir Horowitz traveled with his own piano, but then, Horowitz was Horowitz. You tell him “No.” Unusual among modern pianists, Mitsuko Uchida travels with her own piano. When Tommasini asked her if the institutions where she plays cover that cost, she said “usually not.” But she went on, “I have no excess otherwise. I don’t need country houses, expensive jewelry, expensive cars, special collections of whatever.” I suppose her usual fees cover that cost and still provide her with lunch money.

Tommasini concluded the column: Back at my apartment, the technician finally dropped by, tuned my piano, and made mechanical tweaks to a few of the keys. Afterward, it felt and sounded vastly better. I have no idea what was involved.

Press the key and the pipe blows.

The pipe organ is the most complex of all musical instruments. It is such a sophisticated machine that other musicians, including some world-renowned orchestral conductors, consider it to be unmusical. While a violinist or clarinetist can accent a note by applying a touch more energy, what a single organ pipe can do is all it can do. The organist can accent a note by tweaking the rhythm—a nano-second of delay can translate into an accent—or by operating a machine. A twitch of the ankle on the Swell pedal does it, so does coupling a registration to another keyboard with a soft stop so a note or two can be accented by darting to the other keyboard. The creative organist has a bag of tricks that bypass the mechanics and allow the behemoth to sing.

I have been building, restoring, repairing, servicing, selling, and relocating pipe organs for over forty-five years, and I know that many organists have little idea of how an organ works, so I thought I would offer a short primer. If you already know some or most of this, maybe you can share it with people in your church to help them understand the complexity. In that case, it might help people, especially those on the organ committee, understand why it is so expensive to build, repair, and maintain an organ.

Pipes and registrations

A single organ pipe produces a tone when pressurized air is blown into its toehole. The construction of the pipe is such that the puff of air, which lasts as long as the key is held, is converted to a flat “sheet” that passes across the opening that is the mouth of the pipe. The tone is generated when the sheet is split by the upper lip of the mouth. This is how tone is produced by a recorder, an orchestral flute, or a police whistle. Organ pipes that work this way are called “flue pipes,” and there are no moving parts involved in tone production. Reed pipes (trumpets, oboes, clarinets, tubas, etc.) have a brass tongue that vibrates when air enters the toehole: that vibration is the source of the tone.

Since each pipe can produce only one pitch, you need a set of pipes. We call them ranks of pipes, with one pipe for each note on the keyboard to make a single organ voice. Additional stops are made with additional ranks. There are sixty-one notes on a standard organ keyboard. If the organ has ten stops, there are 610 pipes. Pedal stops usually have thirty-two pipes.

The Arabic numbers on stop knobs or tablets refer to the pitch at which a stop speaks. 8′ indicates unison pitch because the pipe for the lowest note of the keyboard must be eight feet long. 4′ indicates a stop that speaks an octave higher, 2′ is two octaves higher, 16′ is an octave lower. Some stops, such as mixtures, have more than one rank. The number of ranks is usually indicated with a Roman numeral on the stop knob or tablet. A four-rank mixture has four pipes for each note. The organist combines stops of different pitches and different tone colors to form a registration, the term we use to describe a group of stops chosen for a particular piece of music or verse of a hymn.

The length of an organ pipe determines its pitch. On a usual 8′ stop like an Open Diapason, the pipe for low CC is eight feet long, the pipe for tenor c° is four feet, for middle c′ is two feet, and the highest c′′′′ is about three inches. Every organ pipe is equipped with a way to make tiny changes in length. Tuning an organ involves making those tiny adjustments to hundreds or thousands of pipes.

Many organs have combination actions that allow an organist to preset a certain registration and recall it when wanted by pressing a little button between the keyboards (piston) or a larger button near the pedalboard to be operated by the feet (toestud).


When playing a piece of music on an organ, the little puff of air through each organ pipe to create sound is multiplied by the number of notes and the number of stops being used. Play the Doxology, thirty-two four-note chords, on one stop and there will be 128 puffs of air blowing into pipes. Add a single pedal stop to double the bass line and you will play 160 pipes. Play it on ten manual stops and two pedal stops, 1,384. A hundred manual stops (big organ) and ten pedal stops, 6,420, just to play the Doxology, a veritable gale.

Where does all that wind come from? Somewhere in the building there is an electric rotary blower. In smaller organs, the blower might be right inside the organ, in larger organs the blower is typically found in a soundproof room in the basement. The blower is running as long as the organ is turned on, so there needs to be a system to deal with the extra air when the organ is not being played, and to manage the different flow of air for small or large registrations. The wind output of the blower is connected to a unit that most of us refer to as a bellows. “Bellows” actually defines a device that produces a flow of air—think of a fireplace bellows. Before we had electric blowers, it was accurate to refer to the device as a bellows. When connected to a blower that produces the flow of air, the device has two functions, each of which implies a name. It stores pressurized air, so it can accurately be called a reservoir, and it regulates the flow and pressure of the air, so it can accurately be called a regulator. We use both terms interchangeably.

Between the reservoir/regulator and the blower output, there is a regulating valve. Sometimes it is a “curtain valve” with fabric on a roller that operates something like a window shade, and sometimes it is a wooden cone that seats on a big donut of felt and leather to form an air-tight seal. In either case, the valve is connected to the moving top of the reservoir/regulator. When the blower is running and the organ is not being played, the valve is closed so no air enters the reservoir. When the organist starts to play, air leaves the reservoir to blow the pipes, the top of the reservoir dips in response, the valve is pulled open a little, and air flows into the reservoir, replenishing all that is being used to make music by blowing pipes.

Weights or springs on the top of the reservoir regulate the pressure. The organ’s wind pressure is measured using a manometer. Picture a glass tube in the shape of a “U,” twelve inches tall with the legs of the “U” an inch apart. Fill it halfway with water, and the level of the water will be equal in both legs. With a rubber tube, apply the pressure of the organ’s wind, and the level of the water will go down on one side of the “U” and up on the other. Measure the difference and voilà, you have the wind pressure of the organ in inches or millimeters. It is common for the wind pressure to be three inches or so in a modest tracker-action organ. In a larger electro-pneumatic organ, the pressure on the Great might be four inches, six inches on the Swell, five inches in the Choir, with a big Trumpet or Tuba on twelve inches. The State Trumpet at the Cathedral of Saint John the Divine in New York City is on 100 inches. I used to carry a glass tube full of water into an organ, a risky maneuver. Now I have a digital manometer.

In a small organ, the blower typically feeds a single reservoir that regulates the flow and pressure and distributes the wind to the various windchests through wind conductors (pipes), sometimes called wind trunks. In larger organs, it is common to find a regulator in the basement with the blower, and big pipes that carry wind up to the organ where it distributes into various reservoirs, sometimes one for each keyboard or division. Very large organs have two, three, four, or more windchests for each keyboard division, each with its own reservoir. A large bass Pedal stop might have one reservoir for the lowest twelve notes and another for the rest of the stop. And speaking of big pedal stops, the toehole of the lowest note of something like a 16′ Double Open Wood Diapason can be over six inches in diameter. When that valve opens, a hurricane comes out.


The organ’s pipes are mounted on windchests arranged in rows on two axes. All the pipes of one rank or stop are arranged in rows “the long way,” and each note of the keyboard is arranged in rows “the short way.” The keyboard action operates the notes of the windchests, and the stop action determines which sets of pipes are being used. Pull on one stop and play one note, and one pipe plays. Pull on five stops and play a four-note chord, and twenty pipes play. In a tracker-action organ or an electric-action organ with slider chests, the keyboard operates a row of large valves that fill a “note channel” when a note is played and a valve opens. The stops are selected by sliders connected to the stopknobs, which have holes identical to the layout of the holes the pipes are sitting in. When the stop is off, the holes do not line up. When the stop is on, they do, and the air can pass from the note channel into those pipes sitting above open sliders.

It is common in electro-pneumatic organs for there to be an individual valve under every pipe. There is an electric contact under every note on the keyboard, a simple switch that is “on” when the note is played. The current goes to the “primary action” (keyboard action) of the windchest. The stops are selected through various devices that engage or disengage the valves under each set of pipes. When a note is played with no stops drawn, the primary action operates, but no pipe valves open. The stopknobs or tablets have electric contacts similar to those in the keyboards. When a stop is turned on and a note is played, a valve opens, and a pipe speaks.

We refer to “releathering” an organ. We know that the total pipe count in an organ is calculated by the number of stops and number of notes. An organ of average size might have 1,800, 2,500, 3,000 pipes. Larger organs have 8,000 or 10,000 pipes, even over 25,000. The valves under the pipes are made of leather, as are the motors (often called pouches) that operate the valves. Releathering an organ involves dismantling it to remove all the internal actions, scraping off all the old leather, cutting new leather pieces, and gluing the motors and valves in place with exacting accuracy. The material is expensive, but it is the hundreds or thousands of hours of skilled labor that add up quickest.

It’s all about air.

We think of the pipe organ as a keyboard instrument, but that is not really accurate. A piano’s tone is generated by striking a string that is under tension and causing it to vibrate. That is a percussion instrument. The tone of the pipe organ is generated by air, either being split by the upper lip of the organ pipe or causing a reed tongue to vibrate. The organ is a wind instrument. When we play, we are operating machinery that supplies and regulates air, and that controls the valves that allow air to blow into the pipes. When I am playing, I like to think of all those valves flapping open and closed by the thousand. I like to think of those thousands of pipes at the ready and speaking forth when I call on them like a vast choir of Johnny-One-Notes. I like to think of a thousand pounds of wood shutters moving silently when I touch the Swell pedal. I believe my knowledge of how the organ works informs my playing.

A piano is more intimate than a pipe organ, though technically it is also played by remote control as a mechanical system connects the keys to the tone generation. I am not surprised, but I am curious why more pianists do not make a study of what happens inside the instrument when they strike a key. I believe it would inform their playing. A clarinetist certainly knows how his tone is generated, especially when his reed cuts his tongue.

I have always loved being inside an organ when the blower is turned on. You hear a distant stirring, then watch as the reservoirs fill, listen as the pressure builds to its full, and the organ transforms from a bewildering heap of arcane mechanical gear to a living, breathing entity. I have spent thousands of days inside hundreds of organs, and the thrill is still there. 

That’s about 1,800 words on how an organ works. My learned colleagues will no doubt think of a thousand things I left out. I was once engaged to write “Pipe Organs for Dummies” for a group of attorneys studying a complex insurance claim. It was over twenty-five pages and 15,000 words and was still just a brief overview. Reading this, you might not have caught up with Mitsuko Uchida, but you’re miles ahead of Jeremy Denk.

A postscript

In my column in the November 2020 issue of The Diapason (pages 8–9), I mentioned in passing that G. Donald Harrison, the legendary president and tonal director of Aeolian-Skinner, died of a heart attack in 1956 while watching the comedian-pianist Victor Borge on television. The other day, I received a phone message from James Colias, Borge’s longtime personal assistant and manager, wondering where I got the information. I have referred to that story several times and remembered generally that it was reported in Craig Whitney’s marvelous book, All the Stops, published in 2003 by Perseus Book Group. Before returning Colias’s call, I spoke with Craig, who referred me to page 119, and there it was.

I returned Mr. Colias’s call and had a fun conversation. He told me that he had shared my story with Borge’s five children (now in their seventies). He also shared that when Victor Borge was born, his father was sixty-two-years-old, so when he was a young boy, he had lots of elderly relatives. His sense of humor was precocious, and when a family member was ailing, he was sent to cheer them up. Later in life, Borge said that they either got better or died laughing. I guess G. Donald Harrison died laughing.

Photo: Tracker keyboard action under a four-manual console, 1750 Gabler organ, Weingarten, Germany. (photo credit: John Bishop)

Related Content

August 22, 2021
Five manuals? Are you kidding? Most organs around the world have two manual keyboards and a pedalboard. Three manual organs are common, as are “four-…
June 27, 2021
This article is the final installment of a series in the August 2015, June 2016, July 2017, February 2018, June 2018, March 2021, and May 2021 issues…
April 27, 2021
This article is a continuation of a series in the August 2015, June 2016, July 2017, February 2018, June 2018, and March 2021 issues of The Diapason…