|
|
|
Robert
Hope-Jones and the Pipeless Organ Colin
Pykett
Stuart Clark writing of the probable cataclysm were a solar electrical storm to recur like the one which took place in 1859, the year Robert Hope-Jones was born [13]
Posted:
16 February 2010 Last revised: 16 April 2012 Copyright © C E Pykett
Robert Hope-Jones, 1894
Abstract.
Robert Hope-Jones is best known as an innovative organ builder of the
Victorian era in Britain, and subsequently in America. However, his clear
vision of a pipeless organ which he described to the College of Organists in
London (later the RCO) in 1891 is one of the earliest, if not the earliest,
milestones in the codified history of that instrument.
This aspect of his lecture was remarkable because it occurred some two decades
before the first primitive triode valves (vacuum tubes) appeared, which was
about the same time that the word 'electronics' itself was coined. It
is historically important that the transcript of his lecture was published both
by the College and by the musical press, because the date is thereby fixed
unambiguously and because of the level of detail which was revealed.
The latter demonstrates that he had devoted much thought to his ideas,
which arose largely from his background and experience as a telephone engineer.
This article examines the confident claims made by Hope-Jones in his
lecture, and suggests various ways in which the limited electrical technology of
the late Victorian era might have enabled him to realise them in practice.
In particular, the embryonic state of the art in pre-electronic spectrum
(wave) analysis, oscillators, amplifiers, loudspeakers and signal mixing are
discussed in detail in this article. Hope-Jones
never seems to have built a working prototype, that mantle thereby falling a few
years later on Thaddeus Cahill in the USA with his Dynamophone or Telharmonium
which was probably conceived independently.
Nevertheless, his lecture to the College of Organists gives him
indisputable precedence as the probable originator of the idea of a pipeless
organ using additive synthesis. This was twenty years before the slightest vestige of the
appropriate electronic technology became available, nearly half a century before
Hammond and Compton succeeded in realising the technique commercially to a
limited extent, and nearly a century before it was finally implemented digitally
at Bradford university in the 1980's. Therefore,
as a fascinating case study in sophisticated and accurate technical forecasting,
his lecture of 1891 is hard to beat. Contents (click
on the headings below to access the desired section) A
Victorian Additive Synthesiser or 'Electric Organ'
The
origins of today's pipeless organs are surprisingly elusive because they go back
far earlier than the time of the Hammond and Compton instruments of the 1930's.
Perhaps the biggest historical paradox is that people were talking about
pipeless organs a long time before electronics as we know it today was conceived
of, and in fact well before the word 'electronics' itself was coined.
Because not everyone is acquainted with this remarkable fact, it is worth
asking you to read a rather long extract from a lecture which Robert Hope-Jones,
the famous Victorian pipe organ builder, gave to the College of Organists (the
forerunner of the Royal College of Organists) in 1891 [1].
As well as describing his novel electric action which he had already applied to
pipe organs, he also said: "I
should like, if you will permit me, to refer to an instrument which is, I am
convinced, destined sooner or later entirely to supersede the present wind
organ. I
refer to what may rightly be described as the "Electric Organ".
This is a subject of deepest interest, and makes one long for leisure
uninterruptedly to follow it out. We
are not yet able to introduce the electrical organ in a commercial form, yet I
think it may interest you to hear some of the results I have already obtained. An instrument can now be constructed which, though more
expensive, has such advantages over our present organ as the following: it is
actuated solely by electricity, and requires no bellows, soundboard, or pipes.
It is tuned by the maker once and for all, future variation in this
respect being an absolute impossibility. The
tones of the various registers are obtained in a most beautiful and scientific
manner, by combination of the various upper partials with the ground tone, each
in a differing degree. By this
means any kind of timbre can be exactly reproduced, and the tone of each stop is
mathematically the same from the top to the bottom of the compass.
In saying any tone, we must understand not only those which existing
organs, orchestras, and voices have rendered familiar to us, but also an
infinite variety of new tones, some of them most charming.
This
instrument is infinitely more expressive than any wind organ can possibly be,
the combined tones of the full organ being reducible to a whisper at pleasure,
while on the other hand the power of any particular stop, combination, or class
of tone, can be raised to an extent limited only by the power of the instrument.
If so desired, the tones of the full organ, or any portion of it, will
die out altogether. Our chords can
literally "tremble away into silence".
Further than this the tone of solo stops may be gradually varied during
use. Electric
organs do not occupy a hundredth part of the space required for the present wind
instruments. They consist essentially of three parts - the producer of the
currents, the organist's console for controlling the same, and the sounding
apparatus. The latter may be placed
at any distance from the performer, and need in most cases be connected only by
two wires. You will notice I here
say "any distance", but in the case of electro-pneumatic wind organs
"reasonable distance". If
too far away from a wind organ, the loss of time before the sound reaches the
performer proves disagreeable, but with the electric organ proper a small
duplicate sounding apparatus is placed on the console, which faithfully
represents in miniature, at the instant of touching the key, the powerful music
issuing from its large and distant brother". Putting
aside the hyperbole which today's digital organ manufacturers still wallow in
(and which Hope-Jones therefore also invented), he very clearly described an
electric musical instrument using the technique of additive synthesis in which a
periodic (cyclically repeating) sound waveform, such as that corresponding to a
single musical note, can be built up from a number of harmonics, each one having
the correct amplitude or strength. In
his words: "the tones of the various registers are obtained ... by
combination of the various upper partials with the ground tone, each in a
differing degree". Each
harmonic is a pure sine wave, their frequencies all being integer multiples of
the fundamental frequency of the note. Thus
if the note is A above middle C, its fundamental frequency is today standardised
at 440 cycles per second or Hertz (Hz), and its harmonics will then lie at
frequencies of 880 (= 2 x 440), 1320 (= 3 x
440), 1760 (= 4 x 440) Hz and so on.
The general idea is illustrated in Figure 1 by showing how a square wave
arises from a series of odd-numbered harmonics only (square waves do not require
even-numbered ones). Only three
harmonics are shown but the square wave (marked 'sum') can be seen emerging as
they are added together. The more
harmonics which are used in additive synthesis, the greater the fidelity of the
re-created sound. Therefore, with
progressively more harmonics, the synthesised square wave in this example would
get progressively 'squarer' and less smooth in shape.
Figure
1. Illustrating additive synthesis During
the year of this lecture Hope-Jones turned 32, and by then he was a mature and
experienced electrical engineer. He
had recently resigned from his senior position as chief engineer of a telephone
company to devote himself to organ building, and he had already revealed to an
astonished world his famous pipe organ with its novel electric action at St
John's church, Birkenhead. Because
what he said above was published both by the College of Organists [1]
and subsequently by the musical press, it is reasonable to grant him the
appropriate precedence in this field. Although
additive synthesis of itself was not new, as we shall see presently, I know of
nothing predating this which so clearly sets out the principles of additive
synthesis as applied to pipeless organs. Hope-Jones's
admirably clear description of it still forms the basis of some digital
instruments today, in particular those using the technology invented at Bradford
university c.1980 and subsequent modifications of it.
Prior to that were the Hammond organ and the Compton Electrone of the
1930's, both of which used additive synthesis.
Going back further still was Cahill's fantastic Telharmonium, patented in
1897. Thus it is not unreasonable
to grant that Hope-Jones conceived the pipeless organ in 1891 because he almost
certainly described it so clearly before anyone else. Looking
at what he said, he must have understood the following: 1.
Organ pipe sound waveforms can be analysed to discover how many harmonics
they contain, and the amplitude of each one. 2.
By generating sufficient of these harmonics electrically with the
appropriate amplitudes and then adding them together, the original sound will be
re-created. As
remarked above, electronics in its earliest form was years away and therefore
amplifiers and loudspeakers as we know them had yet to appear, even though Hope-Jones referred
to the "producer of the currents" and the "sounding
apparatus" as though he was confident he could construct them.
So had he actually managed to analyse organ pipe waveforms to discover
their harmonic structures? And, if he had, how had he done it? Moreover, what non-electronic engineering means did he
propose for recreating the sounds via the process of additive synthesis - in
other words, how did he propose to actually build his electric organ in the
pre-electronic era? These are the
two key questions which are addressed in this article, and to answer them we
first need to go further back into history.
It is necessary to peer carefully beneath a possible veneer of vagueness
and confusion which Hope-Jones may have introduced deliberately, because
elsewhere in his lecture he made claims relating to his pipe organ actions which
were blatantly untrue. I have
discussed this elsewhere on this website [1], and
regrettably it confirms his distasteful reputation as a self-publicist,
sometimes at the expense of truth. We
therefore need to proceed cautiously in assessing what he said as objectively as
possible.
Starting
with the etymology, it is an interesting coincidence that the term 'electron'
was first applied in about 1891, the year of Hope-Jones's lecture, to the
suspected existence of a fundamental unit of electrical charge by the Irish
physicist G J Stoney. However it
was some years before its reality was confirmed experimentally by the English
scientist J J Thomson in 1897. The
derived word 'electronics' did not appear until about 1910, some two decades
after Hope-Jones's lecture, and several years after the invention of thermionic
valves (vacuum tubes) in which electrons were deliberately generated and
controlled. The first of these was
the diode valve invented in 1904 by another Englishman, Ambrose Fleming.
However electronics as we know it could not really take off until the
subsequent appearance of the triode (a diode with a third grid electrode)
introduced by the charismatic American inventor Lee de Forest a few years later.
The triode was the first and vital enabler of the whole gamut of
functions which we associate with electronics today, such as amplification,
oscillation, feedback, logic switching, etc.
It was followed by other, more complex, types of valve which reigned
unchallenged in electronics for several decades until 1947, when another
invention of even greater importance appeared - the transistor.
This was demonstrated by Bardeen, Brattain and Shockley at Bell Telephone
Labs in the USA, and they subsequently won a Nobel prize for their efforts.
In 1959 the Americans Jack Kilby and Robert Noyce miniaturised both the
transistors themselves and their interconnections to produce the integrated
circuit technology we find all around us today. So
'electronics' did not exist at all when Hope-Jones foresaw the pipeless organ in
the 1890's. He could have had no
knowledge of it whatsoever. The
enabling technology available to him was still restricted to batteries, dynamos,
electromagnets and relays. It is
therefore fascinating that he was able to see the potential of these basic
components for generating music electrically, let alone that he may have used
them to actually construct some elements of an electric organ using additive
synthesis. However, he did not have
to start from a state of zero knowledge because he was able to draw on some
discoveries from an even earlier era, as we shall now see.
As
early as the 1830's we find references in the literature to making music by
electrical means, which was long before Hope-Jones was born in 1859.
This was the era of towering figures such as Michael Faraday in Britain
who is credited with the discovery of electromagnetic induction in 1831. In this
case a varying magnetic field produces an electric current.
This was the reverse of the effect whereby an electric current produces a
magnetic field, discovered somewhat earlier by the Dane Ørsted in 1820 and used
by another Brit, William Sturgeon, in 1823 to produce the first electromagnet. Interestingly, Sturgeon had an organ builder friend in
William Wilkinson of Kendal, and it is said that the pair of them tried to open
the pallet valves of an organ electrically and thereby demonstrate the world's
first direct electric pipe organ action. Unfortunately
the experiment would have failed because there was no reliable source of supply
of the high currents needed. So
in this era we are obviously speaking of matters electrical rather than
electronic, but this does not mean there was any lack of interest in seeking a
relation between electricity and music. On
the contrary, the first half of the nineteenth century was one of frenzied
activity in anything remotely electrical, and even Beethoven was moved to say
"I am electrical by nature".
So it is perhaps unsurprising that a doctor practising in Massachusetts,
Charles Page, noticed that a horseshoe electromagnet produced a brief pinging
sound when the current through its coil was interrupted.
Calling it 'galvanic music', he published a paper describing this effect
in 1837 [2] and correctly divined that the magnet was simply
acting like a tuning fork whose prongs were slightly drawn towards each other by
the magnetic force produced by the current.
When the current was interrupted they sprang apart suddenly and produced
a transient musical sound. His
article became well known and it was followed by others dealing with similar
phenomena. Going
further back still, another very important, if unwitting, contributor to the
electrical music scene was Joseph Fourier, the famous Frenchman whose work
resulted in an independent theoretical thread running parallel to the others in
the tapestry just outlined. Drawing
together some earlier work through the power of his mathematical mind, he is
credited as the first to show in about 1807 that periodic (cyclically repeating)
waves, such as those producing musical tones, could be analysed into a retinue
of harmonics of different strengths (amplitudes).
However he himself was apparently little interested at this time in
music, because his famous paper on what we now call the Fourier Series was
actually motivated by the problems of heat conduction in solids [3].
Nevertheless his work also showed that the process was mathematically
reversible, suggesting that the original tone could be re-synthesised by adding
its constituent harmonics together. Today
we call this additive synthesis in the digital music community, and it was the
process described by Hope-Jones in his lecture as we have seen.
In 1863 the astonishingly gifted German polymath, Hermann von Helmholtz,
firmly connected Fourier's mathematics to the fields of hearing and the
perception of music [4].
His experiments showed that our subjective perception of timbre or tone
colour depends strongly on the harmonic ingredients of the sound.
We
also need to review the developments which led to the invention of the
telephone, partly because they built on the foundations just described, partly
because they foreshadowed an early electric piano, and partly because Hope-Jones
was himself a telephone engineer and he would therefore almost certainly have
been familiar with its technical history. Therefore
he was able to draw on this background when formulating the ideas he outlined in
his lecture. From
today's perspective it is perhaps strange that the earliest efforts to make a
telephone were dominated by attempts to send the individual harmonics of sounds
down multiple separate wires, rather than merely sending the broadband signal
down a single pair as we do today. Originally,
this idea arose from the desire to send several simultaneous Morse telegraph
signals using different frequencies, which is what we still use today in
Frequency Division Multiplex (FDM) data transmission systems - this includes DSL
high speed Internet connections, in which the house telephone can be used at the
same time as the Internet without either interfering with the other.
These early ideas were well illustrated by a lecture which Alexander
Graham Bell, who is generally credited with inventing the telephone, gave to the
Society of Telegraph Engineers in 1877 [5].
Following anatomical research on the human vocal tract and auditory
system undertaken with his father, he had made early progress in determining the
frequency structure of speech. Using
methods similar to those devised by Helmholtz, using tuning forks held close to
the resonant cavities of the mouth,
pharynx and larynx, he established the frequencies involved in vowel sounds. However, Bell went further by proposing that these signals
could then be sent down telegraph wires and thereby transmit speech over long
distances for the first time in human history. Initially,
he proposed a sort of spectrum analyser at the transmitting end of the link
based on multiple tuned steel reeds, each of which would send its own frequency
down a wire when set into vibration by the voice. A similar apparatus at the receiving end would then
approximately reconstitute the spoken sounds, much as a vocoder does today.
However his final and immensely important step was to dispense with the
spectrum analyser once he had discovered that a diaphragm moving near a coil of
wire would simply send a broadband electrical copy of the sound wave impinging
on it down the wire. At the receiving end an identical receiver recreated the
original sounds, if rather faintly. This
became the nucleus of the first practical telephone system before Edison
improved it by inventing the carbon granule microphone.
But the important point to bear in mind here is that the telephone was
originally conceived, essentially, as an additive synthesis device. During
his work Bell also came very close to demonstrating an electric piano.
He envisaged a bank of electrically-maintained tuning forks oscillating
continuously, which we would call analogue tone generators today, with the
oscillatory current from each sent to a distant location via its own key contact
operated by a keyboard. At the
receiving end was an array of electromagnets, each energised by one of the
oscillator circuits, placed in proximity to the appropriate string of a piano
soundboard. In his own words: "Why
should not the depression of a key like that of a piano direct the interrupted
current from any one of these forks, through a telegraph wire, to a series of
electro-magnets operating the strings of a piano or other musical instrument, in
which case a person might play the tuning-fork piano in one place and the music
be audible from the electromagnetic piano in a distant city ? The
more I reflected upon this arrangement the more feasible did it seem to me;
indeed, I saw no reason why the depression of a number of keys at the
tuning-fork end of the circuit should not be followed by the audible production
of a full chord from the piano in the distant city, each tuning-fork affecting
at the receiving end that string of the piano with which it was in unison". However,
we should note at this point that this was not an additive synthesis machine.
It was merely one which enabled the individual strings of a second piano
to sound at a distance from the one being played by the performer.
Nor would it have been able to reproduce the dynamics (soft and loud
effects) of that instrument. It was
therefore little more than an intriguing paper curiosity.
Thus,
by the time Robert Hope-Jones was born in 1859, the foundations of future
electrical musical instruments were being laid. Thanks to Fourier, the harmonic structure of musical tones
was understood in theory and Helmholtz would shortly confirm it experimentally.
Using the same frequency-oriented approach at first, the telephone was
also being developed in parallel and Bell had foreseen a rudimentary electric
piano. All this work was leading
towards the development of two necessary machines - the first which could
determine the harmonic structure of a particular tone in terms of the number and
strengths of its constituent harmonics (a spectrum or wave analyser), and
another which could reconstruct that tone from those harmonics (an additive
synthesiser). By the time he was in
his thirties, Hope-Jones was sufficiently far-sighted to identify a further
application for these ideas - a fully electric pipeless organ using the same
principles of additive synthesis. Therefore,
although he did not invent all of the basic technology, it is reasonable to
credit him as probably the first to perceive the application of it which
ultimately led to the electronic organs of the twentieth century.
A
spectrum or wave analyser is the first essential machine without which one
cannot realise a musical instrument which is intended to copy the sounds of
existing ones, such as the pipe organ, using additive synthesis.
It is used to determine how many harmonics are present in a given tone
emitted by an organ pipe, and the amplitude (strength) of each one.
Without this, it is not possible to accurately re-create the sound
subsequently. Why was Hope-Jones
apparently so confident that he could achieve this in pre-electronic days? Although the details are sketchy, we have enough clues to
enable us to draw some conclusions as to how he might have done it.
An interesting account of some of his work in this area is worth
repeating here: "Imagine,
if you can, entering one of the famous old cathedrals on the Continent, in which
the organ is noted for one or two particularly rare tones and by means of a
little instrument, photographing those particular tones.
Then to stretch your imagination a bit further, conceive of taking the
negatives home with you to America and in your laboratory reproducing that tone
identically. This is precisely what
Hope-Jones has done and is doing. This
instrument is necessarily of his own invention and consists of a lead funnel, in
the apex of which is suspended a silvered aluminum mirror.
As sound waves enter the funnel, the mirror vibrates and a needle-point
ray of electric light is reflected on a screen some ten feet away. The movement of this tiny light spot is recorded by a camera
and from the photographic prints, the inventor reproduces the tone in his
workshop. He says that his
scientific tone language is a wonderful thing and its accuracy is absolute.
Once you have learned the language, it tells you all about sound"
[6]. Although
this appeared after Hope-Jones had emigrated to America in 1903, and thus after
others had entered the field, it is possible it described work he had started in
Britain and obviously continued afterwards.
While technology has moved on, the procedure itself is remarkably similar
to that which today's manufacturers of digital organs undertake when producing
sample sets from cathedral organs for their instruments! Although
Hope-Jones may well have built the device described in the extract above for his
own use, he was able to draw on the designs of several similar instruments which
had been in use for several decades. In
his article from the 1870's already referred to [5], Bell
described the use of the 'phonautograph' which traced sound waveforms (in his
case those of the human voice) mechanically on smoked glass plates.
He included some examples of the traces thus obtained.
Somewhat later, William Duddell invented his oscillograph, an early form
of today's familiar oscilloscope, in which the trace was recorded
photographically, as in Hope-Jones's work.
However the mirror, part of a moving coil galvanometer in this case, was
deflected by an electric current rather than mechanically by sound waves. Duddell's instrument was widely used from the outset, for
example by Marconi in his early radio research in the 1900's [7].
(Duddell was also famous for another discovery, his 'singing arc', in
which an arc lamp could be persuaded to emit musical tones, and he demonstrated
it widely. Hope-Jones may well have
known of this). But
having displayed the sound waveforms photographically, how did Hope-Jones
proceed to analyse them to discover how many harmonics they contained, and the
amplitudes of each? In other words,
how did he perform a spectrum analysis? If
the extract quoted above is to be believed, he must have done it because it
mentions "reproducing that tone identically".
In fact the processes involved, though laborious, were already well known
at the time and therefore they were relatively unremarkable.
By making simple measurements with a ruler on a single cycle of the
recorded waveform from its photographic trace, Fourier's much earlier theory
enabled its harmonic structure to be determined.
Until the advent of computers in the mid-twentieth century, scientists
and engineers routinely implemented this manual process, and Figure 2 shows a
typical handwritten 'spreadsheet' designed to facilitate harmonic analysis.
This example was drawn up in about 1930 by Winston Kock who designed the
original Baldwin electronic organ [8].
(So tedious were the necessary calculations in those days that much
effort was devoted to speeding them up, and there is persuasive evidence that
Cooley and Tukey's 'Fast Fourier Transform' computer algorithm of the 1960's was
in fact in routine use several decades earlier). Figure
2. One of Winston Kock's harmonic
analysis 'spreadsheets' [8] A
Victorian Additive Synthesiser or 'Electric Organ'
So
Hope-Jones, in common with anyone else having the interest and inclination at
the time, had access to the necessary technology and know-how to enable him to
discover the harmonic structures of organ pipe sounds.
But could he have built an additive synthesis machine to re-create those
sounds? In other words, could he
have actually built the 'electric organ' he described so vividly in his 1891
lecture to the College of Organists? This
was much more challenging. Such
an organ would have required a set of oscillators producing enough sine waves
for the entire instrument. The
electrical signals from these would be switched in and out of circuit via key
contacts, or via relays operated by them, and they would then need to be
combined (mixed) in the proportions suggested by the spectrum analyses already
carried out. Finally the output
would need to be made powerful enough to operate some form of loudspeaker. But, remember, none of this could call on electronics - it
would all have had to be done using the much more basic technology available in
the 1890's. Several
types of oscillator were available to Hope-Jones, and he would have had to
decide whether they were to generate small signals which would need to be
amplified later in some way, or whether they would generate enough power in the
first place to avoid the need for subsequent amplification.
This was an important issue given that electronics, and therefore the
ability to amplify small signals easily, was years into the future.
Another important factor was the practicality of constructing large
numbers of oscillators, typically 97 being necessary to construct the most basic
instrument if it was to sound anything like realistic.
97 oscillators would cover a frequency range of eight octaves from about
30 Hz (16 foot bottom C) to about 8 kHz (2 foot top C).
Each oscillator would need to generate something close to a sine wave,
and they would have had to be very stable in frequency if Hope-Jones's claim was
to be vindicated that the instrument "is tuned by the maker once and for
all, future variation in this respect being an absolute impossibility". The
options open to him for stable small-signal oscillators would have included an
array of electrically maintained tuning forks as used already by people such as
Helmholtz and Bell. Such a fork
operates on the principle of an electric bell, with an interruptor contact
arranged so that when the tines of the fork bent slightly under the influence of
an electromagnet, the circuit would be broken. The tines would then spring back again and the cycle would
repeat for as long as power was applied. A
sketch of such a fork, based on a design actually used in the mid-nineteenth
century and still produced today for teaching elementary physics, is shown in
Figure 3. Figure 4 is a picture of
an actual nineteenth century fork made by Rudolph König which is identical,
except that the interruptor used a mercury cup and dipping contact rather than a
mechanical contact. Helmholtz used
similar forks in his own work on phonetics and musical acoustics.
Figure
3. Principle of the electrically
maintained tuning fork
Figure
4. An electrically maintained
tuning fork by König (Copyright
© Prof T B Greenslade Jr) [9] The
musical signal could be taken from this system in two ways - either it could
simply be the interrupted current itself, or it could be induced in a separate
coil placed near the tines (e.g. between them). A signal derived from the interrupted current would have had
the advantage of being stronger than an induced one, but it would have had a
pulse-type waveshape rather than being sinusoidal, and therefore it would have
required low pass filtering. Unless
the generated signals were close to sinusoidal, they could not have been used
directly for additive synthesis because the resulting tones would have suffered
from considerable spectral (harmonic) distortion. Filtering was another operation which would have been
difficult at the time.
Figure 5. Electrically maintained tuning fork using a carbon granule microphone
Electrically maintained tuning forks remained under serious consideration for musical tone generation until the middle of the twentieth century, and they were chosen for use in one of the first electronic synthesisers by Harry F Olson, a well known engineer of the time at RCA. Figure 5 shows yet another form of the system in which the mechanical interruptor in the two previous embodiments is replaced by a carbon granule microphone acting as a continuously variable resistance element. This diagram appeared in a book by a well-known electronic music specialist of the day, who also said that "forks are now [1957] available up to at least 10 kc/s" [11]. This is interesting for several reasons. Firstly, the carbon microphone had been invented by Edison specifically for telephone purposes in 1877 [10], thus Hope-Jones (a telephone engineer) would certainly have known about it by the time he gave his 1891 lecture. Secondly, although Figure 5 shows a separate signal coil arranged to pick off a sinusoidal music signal as mentioned above for the mechanically-interrupted fork, the driving current for the fork itself would be much smoother and closer to sinusoidal than the pulse waveform generated by a mechanical interruptor. Because it would also be significantly stronger than that induced in the separate pickup coil, it would have been more attractive on both these counts for use in pre-electronic times. Thirdly, the mention of forks resonating at frequencies up to 10 kHz suggests that Hope-Jones would probably have had little difficulty in achieving a reasonably high frequency response for his proposed electric organ. On balance, therefore, there is strong collateral evidence suggesting that Hope-Jones may have had in mind an array of electrically-maintained tuning forks as a set of small-signal quasi-sine wave generators.
Nevertheless,
another
possibility for small-signal oscillators is suggested by one of Hope-Jones's own
patents. British patent number
15245 (1890) contains an interesting illustration which is reproduced here in
Figure 6. At first glance one might
assume it is an electromagnetic tone generator of the type used much later in
the Hammond organ, because it consists of a rotating metal disc with what is
apparently a pickup coil at the periphery.
The lower sketch in Figure 6 even suggests that the edge of the disc is
serrated or toothed, as in the Hammond instruments.
In fact the description in the patent is of an amplifier, not a signal
generator, and it was apparently intended to amplify weak signals such as those
from a telephone line so they could be heard more easily.
Thus the coil would have been fed from the line, and the amplifying
action occurred due to a combination of electromagnetic and mechanical effects -
the incoming signal resulted in varying amounts of magnetic drag on the disc,
with the consequential (mechanically amplified) vibrations being transmitted
mechanically to the large-area diaphragm L.
The bearings of the disc in the trunnion were presumably pre-loaded with
a certain amount of friction to enable this to occur.
Figure
6. Hope-Jones's rotating disc
amplifier and possible tone generator It
is more than likely that Hope-Jones, either accidentally or by design, might
have seen the possibilities for tone generation using this arrangement.
If he happened to be listening to the signals in the incoming circuit
using an ordinary telephone earpiece, perhaps during testing, it is conceivable
that a faint background of musical tones might also have been heard due to the
rotating disc inducing signals in the coil.
In the patent he recommended that the disc "preferably" be of
copper, with the drag then arising from eddy currents induced in it by the
receiver coil. However, he might have also tried iron or steel with the
intention of increasing the drag forces. In
this case, any residual magnetic inhomogeneities in the disc could have resulted
in stronger tones being generated which might have been audible.
Even without such serendipitous intervention, it is somewhat unlikely
that Hope-Jones, a professional electrical engineer, could have failed to see
the possibilities just outlined for tone generation.
The postulated arrangement is nothing more than a simple low power
alternator whose potential would surely not have been lost on him, just as it
was not lost on the clockmaker Laurens Hammond several decades later. It
was mentioned above that Hope-Jones could also have chosen to use a set of
high-power alternators to avoid the need for subsequent amplification, though it
is difficult to say whether he considered this option or not.
However, he said in his 1891 lecture that "electric organs do not
occupy a hundredth part of the space required for the present wind instruments",
implying that the bank of oscillators he envisaged would have been small and
compact. Had he contemplated the
use of a large number of high-power alternators he could not have made this
statement so confidently. Because he had already designed small-signal amplifiers for
use in telephone systems, such as that above, it seems reasonable to conclude
that he would have been disposed to go down this route, and therefore that he
probably focused his attention on the technique of small-signal generation with
subsequent amplification. We shall
now look at how he might have achieved the latter. Hope-Jones's
rotating disc amplifier has already been mentioned. Among several other systems he described in his patents was a
combined fluidic amplifier and loudspeaker system (patent number 15245/1890),
illustrated here in Figure 7.
Figure
7. Hope-Jones's fluidic amplifier
and loudspeaker system The
illustration is of a device intended as a sort of megaphone.
One would speak into the mouthpiece in the centre of the diagram and an
amplified version of one's voice would be projected from the loudspeaker on the
right hand side. The potential
energy to achieve the amplification was derived from water under pressure, which
flowed continuously through multiple orifices in the valve assembly illustrated
in more detail in Figure 8. By
means which are not entirely clear, the movement of the input diaphragm was
supposed to translate into transverse movements of the valve face closer to and
away from a plate attached to the loudspeaker diaphragm, mechanical
amplification being obtained by virtue of the energy in the fluid stream.
In fact there are errors in the two drawings which are consequently
mutually inconsistent, though one can approximately discern the intended
principle of operation nevertheless. (Hope-Jones's
patents are littered with errors, and one sometimes wonders whether he might
have intended this deliberately. The
problem also arises in some of his patents dealing with circuits for pipe organ
electric actions).
Figure
8. Hope-Jones's fluidic amplifier:
the valve assembly If
the input diaphragm in Figure 7 was replaced by an electromagnetic actuator
which was supplied with the output current from the organ, one could imagine
that the same type of fluidic amplifier and loudspeaker would then project the
sound of the instrument into the auditorium. The
patent describing all the amplifying and loudspeaking devices discussed here
(and several others also included in it) was dated only the year before
Hope-Jones gave his 1891 lecture to the College of Organists.
It is therefore reasonable to presume that he had in mind a system of
this general nature when he mentioned the "sounding apparatus" which
he referred to, though it is unclear how far, if at all, he had gone in actually
building prototypes. However this
work must have arisen largely during his time as a telephone engineer, the
occupation from which he had only just resigned.
Louder telephones were certainly a crying need at the time, Thomas Edison
being just one inventor who had already devoted much effort to the problem (in
his case, mainly to circumvent Bell's patents).
One of the reasons for this was that the receiver was (literally!) the
weakest part of Bell's telephone system from the user's point of view - if used
on long circuits the received signals were barely audible. As early as 1879 Edison had demonstrated his
'electromotograph' telephone receiver which was apparently "of such
stentorian efficiency that it bellowed your most private communications all over
the house instead of whispering them with some sort of discretion" [10].
If you think Hope-Jones's fluidic system was barely credible, then you
would probably conclude at first acquaintance that Edison's apparently bizarre
electromotograph idea was little more than nonsense.
But, even today, it is a brave mortal who would posthumously challenge
Edison without having done the necessary homework, and in this case the homework
shows that his idea worked. Indeed,
it worked so well that it had already been shipped to Britain for use at a
demonstration in the vast spaces of the Crystal Palace in London, an event which
Hope-Jones the telephone engineer may have known about and which he might well
have attended. Therefore, overall,
there can be little doubt that he spoke with some justified confidence when
claiming during his 1891 lecture that the electrical reproduction of loud sounds
could indeed be achieved using the limited technology of the day. We
have now reviewed some of the feasible methods which Hope-Jones could have used
in an additive synthesis instrument to generate the basic signals (sine waves),
and then to amplify and radiate them into an auditorium as sounds.
However, there still exists a missing link which in effect sits between
the two processes. This link is the essential requirement to combine the various
harmonics making up the sound of a given organ stop before it is applied as a
complex, summed, waveform to the amplifier and loudspeaker system.
This mixing or summing process lies at the mathematical heart of additive
synthesis. It
might seem a trivial problem to combine, that is, to add, a number of sine waves
in the desired proportions. After
all, even in Hope-Jones's day, resistors (usually wire wound ones) were in
common use in electrical circuits and he mentioned their use in several of his
patents. Therefore each harmonic
could be represented easily as a particular value of current defined by the
appropriate value of resistance in some sort of combining or mixing circuit.
But herein lies the problem - how would the mixing, the actual addition
of the several currents, have been done accurately and efficiently?
What would that circuit actually have looked like?
We need to recall that, strangely enough, this problem was not solved
satisfactorily in electronics until the advent of operational amplifier
techniques, originally developed for analogue computers during the second world
war but not brought into common usage until the 1960's when op-amps in
integrated circuit form became widely available for the first time.
Figure
9. Mixing analogue signals using an
operational amplifier This
modern method is illustrated in Figure 9. As
many input signals as desired are fed into a set of input resistors which are
all connected to the inverting input of an op-amp, shown by the triangular
symbol. (In this case the word
'inverting' means 'multiplied by -1'). Each
resistor is chosen to define the proportion of its input signal which is desired
at the output (the higher the resistance, the smaller the proportion).
The voltage at the output of the amplifier is then the (inverted)
weighted sum of all the input signals, multiplied by a fixed gain defined by the
feedback resistor. The advantage of
this circuit is twofold - firstly, the inputs do not mutually interact, in other
words one can vary the input resistor for any input without it affecting any
other. Therefore the addition is
accurate in a mathematical sense. Secondly,
there is no overall loss of signal strength because the gain resistor can be
chosen to provide any degree of amplification desired within wide limits. Therefore the process is efficient in practice, in the sense
that it compensates for the energy lost as heat in the various resistors. Before
integrated circuit op-amps became available the same function was performed
using discrete transistors or, prior to that, with valves (vacuum tubes).
However in Hope-Jones's day nothing like this was possible because, as we
have seen, electronics was years away - the first primitive amplifying valves
did not become available until around 1910.
Therefore he had only two means at his disposal for mixing signals in the
1890's. The first was to use a simple resistor network as shown in
Figure 10. Here, the various input
signals are applied to a set of resistors as before, but the junction of these
is simply terminated at a common shunt resistor labelled R in the diagram.
The output is also taken from this junction.
This circuit will work, up to a point, but it has two major
disadvantages. Firstly, the various
inputs are no longer functionally isolated from each other as they were in the
op-amp circuit of Figure 9. This
means that if one wanted to vary the strength of a particular harmonic in the
mixed signal, which one would commonly wish to do when voicing the instrument,
one could not simply vary its associated input resistor.
Doing this would also affect the strengths of all the other harmonics as
well. However this problem can be
alleviated by reducing the value of the shunt resistor R.
Unfortunately this then results in the second disadvantage of this
circuit - a lower value of R means that the amplitude of the combined output
signal gets correspondingly lower. Thus
the circuit wastes energy which is converted into heat in the resistors, and
without an amplifier this lost energy cannot be compensated for.
Hope-Jones could not have tolerated a system which threw signal energy
away because, without any means of regaining it easily, he needed to retain all
the power he could to eventually operate the primitive amplifiers driving his
equally primitive loudspeakers.
Figure
10. Mixing analogue signals using a
resistor network There
is little doubt that Hope-Jones would therefore have used a second mixing
method. Because the signals are all
in the form of alternating rather than direct currents, he would probably have
mixed them using an audio frequency transformer. This technique was already in widespread use in telephony in
his day and he would therefore have been familiar with it.
The method is shown in Figure 11.
Figure
11. Signal mixing using a
transformer Three
signal inputs are shown for illustrative purposes, each connected to a separate
primary winding on a transformer with a laminated iron core, though any
reasonable number could be used. The
resistor R in series with each input would enable voicing to be carried out by
adjusting the levels of each harmonic independently of the others - because of
the action of the transformer they scarcely interact provided the system is
properly designed for the purpose.
The mixed output, representing the summed input signals, is then taken
from the secondary winding.
By adjusting the relative number of turns on the primary and secondary windings,
the transformer would also enable the output impedance of the generator system
and the input impedance of the following stage (probably an amplifier of some
sort) to be matched. This would maximise the transfer of power between the
two and thus prevent energy being wasted as discussed already. As
just stated, these matters were well understood in the late nineteenth century.
Transformer
mixing and impedance matching of multiple audio signals
remained the standard method for many years,
and it was used in the first Hammond organs in the 1930's for combining
the nine harmonics from the drawbars.
Having
described Hope-Jones's vision of a pipeless organ using additive synthesis,
together with the means at his disposal for realising it, we now need to briefly
refer to a similar instrument which was actually built only a few years later.
Full details need not be given because they are widely available
elsewhere, including on the Internet. In
1897, only six years after Hope-Jones's lecture to the College of Organists,
Thaddeus Cahill in the USA patented, and then built, a similar system to that
foreseen by Hope-Jones. Initially
called the Dynamophone but later the Telharmonium, it used similar methods for
generating the harmonics and mixing them to those which have just been outlined.
The main difference between Cahill's implementation and that which
Hope-Jones probably envisaged was that Cahill used a large-signal approach
rather than a small-signal one. Therefore
he did not need to bother about amplification which was, on the face of it, a
major simplification. However it
meant that Cahill's system was very large and clumsy because he used many
relatively high-power alternators to generate the various harmonic frequencies.
He used transformers for mixing and tuned chokes for filtering, and his
loudspeaker was nothing more than a telephone-type earpiece mounted at the apex
of a large horn, driven by currents measured in ampéres generated directly by
the alternators. Hope-Jones
became familiar with the Telharmonium after he had emigrated to America and he
obviously tried it because he wrote, rather thoughtfully, that "it is
interesting to observe how the resultant C (fundamental) is still heard when the
ground tone has been practically shut off and nothing but a chord of harmonics
is left sounding" [12].
This interesting psycho-acoustic property of the human auditory system
has been verified many times since. It
is interesting to speculate whether Hope-Jones and Cahill proceeded
independently at almost the same time with their researches on the two sides of
the Atlantic, or whether there was some degree of collaboration or even
plagiarism involved in one direction or the other.
It is likely that there was not, as the plain fact is that Hope-Jones
became involved far more rapidly than he probably anticipated with his pipe
organ building career. When he gave
his 1891 lecture to the College of Organists he had only built a single pipe
organ at St John's, Birkenhead and, although this was already a well known
instrument, he could have had little inkling of how busy, famous and successful
he was to become as an organ builder of almost instant worldwide renown. Although his lecture demonstrates that he had obviously gone
a long way in designing his pipeless organ on paper, he had almost certainly not
been able to develop it in the sense of building a working prototype at that
time. His own words reveal this
when he said that "this is a subject of deepest interest, and makes one
long for leisure uninterruptedly to follow it out", and "we are
not yet able to introduce the electrical organ in a commercial form, yet I think
it may interest you to hear some of the results I have already obtained". These
remarks imply that he had yet to build more than a few components or sub-systems
of it at best, perhaps based on the ideas discussed in this article.
Nor (unusually for him) had he got round to patenting it.
Moreover, shortly afterwards he became so frenetically busy with his pipe
organ work (only five years later he had completed his monstrous four manual magnum
opus at Worcester cathedral) that one assumes he was simply unable to find
the time to do little more for the remainder of his tragically short life than
tinker with the pipeless organ. Given
that there is little or no evidence of acrimony or conflict, he therefore seemed
content to allow Cahill to adopt the mantle of the builder of the world's first
additive synthesis organ, even though he had probably conceived it somewhat
earlier and independently of Cahill. Cahill's
endeavours also reflected the penalty of being a pioneer in that the several
undesirable aspects of the Telharmonium materially hastened its end after only a
few years. Witnessing this
unfortunate story from his vantage point as one of America's most famous pipe
organ builders, Hope-Jones may well have decided that discretion was the better
part of valour as far as additive synthesis was concerned.
Hope-Jones's
unequivocal description of a pipeless organ using additive synthesis which he gave to the College of
Organists in 1891 is one of the earliest, if not the earliest, milestone in the
history of that instrument. It is
historically important that the transcript of his lecture was published both by
the College and by the musical press, because it fixes the date
unambiguously and because of the level of detail which was revealed.
The latter, together with the related material in his patents,
demonstrates that he had devoted much thought to his ideas, which arose largely
from his background and experience as a telephone engineer.
This article shows that Hope-Jones would most likely have proceeded using a set of small-signal tone generators producing quasi-sine waves which would have required subsequent amplification before they would have been powerful enough to operate some form of loudspeaker. This assumption, rather than one which involved high-power signal generation from the outset, is more consistent with his background as a telephone engineer accustomed to handling audio frequency signals of very low power. It is entirely possible that he envisaged the use of a set of electrically-maintained tuning forks as the tone sources, given that they were routinely used in his day by researchers in other fields, and that they remained in use for electrically-produced music until well into the twentieth century. However there is persuasive collateral evidence that he might also have considered rotating electromagnetic tone generators of the type subsequently used in the Hammond organ. He would have had little option than to use transformers for mixing the various harmonics, again a technique which was not displaced in electronic engineering until the mid-twentieth century. Hope-Jones described in his patents several options for non-electronic amplifying and loudspeaking equipment, some of which were similar in principle to those invented by others such as Edison. Thus, overall, although an instrument built along these lines would appear bizarre with today's hindsight, there can be no argument but that it would have worked after a fashion because it was all based on proven technology of the day. Hope-Jones
never seems to have actually built a working prototype, that mantle thereby
falling a few years later on Thaddeus Cahill in the USA with his Dynamophone or
Telharmonium which was probably conceived independently.
Cahill's achievement probably did not result in Hope-Jones discarding the
concepts he had described or losing interest in them, but the fact he became so
frantically busy as a pipe organ builder meant, quite simply, that he was
thereafter unable to devote the necessary time to developing his vision of a
pipeless organ. Nevertheless, his
lecture gives him indisputable precedence as the probable originator of the idea
of a pipeless organ using additive synthesis.
This was some two decades before a vestige of the appropriate electronic
technology became available, almost half a century before Hammond and Compton succeeded
in realising the technique to a limited extent, and nearly a century before it
was finally implemented digitally at Bradford university in the 1980's. Therefore, as a case study in sophisticated and accurate
technical forecasting, his lecture of 1891 is difficult to beat.
Finally we might reflect that, although Hope-Jones is often and rightly called the "father of the theatre organ", it would not be unreasonable to credit him with a contribution to the paternity of the electronic organ also. I
am grateful to Professor Thomas B Greenslade Jr for permission to use the
illustration of König's electrically maintained tuning fork from his website [9]. 1.
"Electrical Aid to the Organist", R Hope-Jones, Proceedings
of the College of Organists, 5 May 1891.
2.
Benjamin Silliman's Journal, (later The American Journal of
Science), C G Page, April 1837. 3.
"Mémoire sur la propagation de la chaleur dans les corps solides,
présenté le 21 Décembre 1807 à l'Institut National, Paris", Nouveau
Bulletin des Sciences par la Société Philomatique de Paris, March 1808.
4.
"On the Sensations of Tone", a translation by Ellis in 1872 of
the German original by H L F von Helmholtz, Brunswick, 1865. 5.
"Researches in Electric Telephony by Professor Alexander Graham
Bell", The Society of Telegraph Engineers, London, 31 October 1877. 6.
"The Master Organ Builder, Robert Hope-Jones and his Unit
Orchestra", A L Miller, Popular Electricity, May 1913 (quoted in
"Robert Hope-Jones", D H Fox, Organ Historical Society, Richmond,
Virginia, 1992). 7.
"The Principles of Electric Wave Telegraphy", J A Fleming,
Longmans, Green & Co, London, 1906. 8.
"Winston Kock and the Baldwin Organ", C E Pykett, 2008.
Currently on this website (read). 9.
See http://physics.kenyon.edu/EarlyApparatus/index.html
(Accessed 22 January 2010) 10.
"Edison", Ronald W Clark, Macdonald and Janes, London, 1977.
11. "The Electrical Production of Music", Alan Douglas, Macdonald, London, 1957. 12.
"The Future of the Church Organ", Robert Hope-Jones, The New Music
Review, February 1908.
13. "The Sun Kings: The Unexpected Tragedy of Richard Carrington and the Tale of How Modern Astronomy Began", Stuart Clark, Princeton 2007.
|
