Reger and the Performance Practice of his Era - the Welte recordings,

including Reger’s own, from the collection at the Museum für

Musikautomaten, Seewen (Switzerland).

David Rumsey

Paper given at the International Reger Symposium

Bruckner Univsersity, Linz, Austria
Wednesday 13th April 2005.

The great International Exhibitions of the 19th and early 20th centuries
were events in which the organ played a "role(roll)" in the fullest
sense of the word. The Welte company’s "Orchestrion"
and"Vorsetzer" were familiar sights. Welte's "Cabinet player", a
reproducing piano without keyboard which bore the Mignon label,
was first patented in 1904 while the firm was under the direction of
Edwin Welte (1876-1958, grandson of the founder). The prototype
was exhibited during late 1904 in Leipzig and became commercially
available from early 1905. The Vorsetzer came on the market in
1908. Mignon was integrated into their upright pianos in 1909, and Michael Welte 1807-1880
into their grand pianos from 1913. In 1908 this technology was also
applied to the organ when the firm unveiled their "Welte-Philharmonie-Autograph-Orgel". It
became the prototype for their "Welte-Philharmonie” player organ1.

In the Welte premises shown here2 - with its workshops, Recording Hall, "House Organists",
etc. - many famous artists of the era were engaged to
record organ rolls for later sale. Amongst them were
names such as Gigout, Straube, Ramin, Goss-Custard,
Wolstoneholme and Reger. Between c1911 and 1928
well over 2000 rolls were recorded; exactly how many
we may never know3.
Welte Premises Freiburg
According to the manufacturer
their system could faithfully reproduce every minute original detail of
a recorded performance. Welte’s success with player organs can be
dated back to the Turin exhibition of 1911. They were soon in great
demand and were available in a variety of sizes and models.

Welte’s assurances of playback fidelity in these reproduced

performances was eagerly taken up by a generally enthralled group of
customers and artists. There can be no doubt that this was a gigantic
achievement in both technical and musical worlds.

The Museum for Automatic Musical Instruments in Seewen

(Switzerland, near Basel) owns a large Welte-Philharmonie organ. There were never many of
this model. For a Welte the Seewen instrument is unusually complete and today world-wide
a great rarity, since it is fully original and furnished with a keyboard so that it can also be
played by an organist, just like any normal organ of its era.
 2

The specifications of Seewen and some other Welte organs may be compared in the
Appendix below.

Most of the organ rolls sold by Welte are also

today preserved in copies or mother-rolls at
Seewen. Many of them originated as the
property of Werner Bosch, a former employee of
the Welte company who acquired them when the
firm was liquidated4. They were, in turn, passed
on in 1970 by Bosch to Heinrich Weiss, the
founder of the Seewen Museum. This collection
of some 1300 organ rolls is now carefully stored
in Seewen’s archives. Almost all remain
The organ installed at Seewen (swell shutters removed) completely intact.
prior to the rebuilding works at the Museum
In about 1980 Swiss Radio DRS featured a
series of programs from the Seewen recordings. I transferred the 5 radio tapes to computer
CDs in November 2004 in order to preserve them. After a quarter of a century the original
recordings were showing clear signs of deterioration. The following shows the detail of what
was recorded and by whom. Organists are shown on the left and the abbreviated title (as
entered in the original Seewen catalogue) on the right.

Binninger Reger Toccata a Op 80 Nr 11

Bonnet Bach Praeludium und Fuge e BWV 533
Bonnet Bonnet Angelus du Soir
Breitenbach Bach Fuge Es BWV 552
Dupré Bach Praeludium und Fuge e BWV 548
Eddy Reger Pastorale op59 nr2
Eddy Liszt BACH
Fischer Reger aus Sonata 2 Opus 60 Invocation
Fischer Reger Gloria Op 59 Nr 8
Fischer Liszt Weinen Klagen
Fischer Bach Triosonata Es dur
Gigout Bach Praeludium Es BWV 552
Goss Custard Bach Fuge Es BWV 552
Grosse Widor Organ Symphony V Op 42 Satz 3
Grosse Reger Toccata und Anfang der Fuga D
Grosse Widor Sinfonie V Op 42 Satz 1
Grosse Reger BACH op 46
Grosse Widor Sinfonie V Op 42 Satz 4
Haeuser Heuberger Opernball
Hindermann Boellman Suite Gothique
Hindermann Reger Ave Maria Opus 63 nr 7 and 80 nr 5
Hofmiller Widor Sinfonie Andante Cantabile
Lemare Lemare Improvisation
Mania Franck Choral Nr 2 h moll
Matthaei Bach Orgelbuechlein Nun komm
 3

Matthaei Bach Orgelbuechlein Christe du Lamm Gottes

Matthaei Bach Orgelbuechlein Wenn wir in höchsten Nöten sein
Matthaei Bach Orgelbuechlein Es ist das Heil uns
Matthaei Pachelbel Ciacona f
Matthaei Sweelinck Mein junges Leben hat ein End
Messner Reger Romanze a Op 80 # 8
Messner Reger Consolation op65 nr4
Nater Guilmant Verset
Nater Widor Sinfonie V Satz 2
Nater Widor Sinfonie V Satz 5
Nater Dubois Noel
Nikisch Brahms Ungarischer Tanz 5
Philipp Reger "Weihnachten" Op 145 Nr 3
Ramin Reger Toccata und Fuga d Op 129 1 und 2
Ramin Bach Partita Sei Gegruesset Jesu guetig
Ramin Buxtehude Praeludium g
Reger Reger Melodia op59 nr11
Reger Reger Benedictus op59 nr9
Reger Reger Romanze g Op 69 Nr 8
Reger Reger Fuge G op56 nr3
Reger Reger Praeludium F Op 85 Nr 3
Reger Reger Canzona op65 nr9
Reger Reger Suite Op 92 Basso Ostinato
Reger Reger Moment Musical D Op 69 Nr 4
Reger Reger Jesus meine Zuversich op67 nr20
Reger Reger Lobt Gott op67 nr23
Reger Reger O Welt ich muss dich lasse op67 nr33
Reger Reger O wie selig op67 nr52
Reger Reger Wer nur den lieben Gott op67 nr45
Reger Reger Wie wohl op67 nr50
Sittard Reger Benedictus Op 59 Nr9
Sittard Bach Praeludium e BWV 533
Sittard Sittard Choralstudie
Stark Reger Vater Unser Op 67 Nr 39
Straube Bach Fantasia g 542
Straube Orgelbuechlein Wir danken dir Herr Jesus Christ
Straube Buxtehude Praeludium g
Straube Orgelbuechlein Christ lag in Todesbanden
Straube Orgelbuechlein Christe, du Lamm Gottes
Straube Orgelbuechlein Da Jesus an dem Kreuze stund

oooOOOooo

In spite of Welte’s advertising hype there were some small flaws with their system.

The narrow width of Welte rolls were problematic. With a mere 150 holes available it was
difficult to record all the notes from 2 manuals and pedals - alone over 140 keys - alongside
 4

the more than 30 stops, accessories, crescendo swells etc.. The system therefore operated
what we might now call a multiplex technology (in this case a pneumatic multiplex). To gain
the extra functiions, for instance, when a certain accessory was operated it might cause two
mutually opposing signals - e.g. the manual I Bordun 16’ is operated when both “slow open”
and “slow close” occur simultaneously.

This was the reason that the pedal was recorded with the main manual on a single
multiplexed track - on replay a pneumatic switching device had to sort out the low notes and
channel them through to play on the correct division. However during replay, it took a
noticeable moment of time to operate this pneumatic switch and decide which of the 30
lowest notes belonged to the Pedal and which to the main manual. Mechanical intervention
in moving notes slightly forward or back was a normal part of the editing process undertaken
by technicians at Welte: these manual editing processes typically extended some of the
original perforations made during recording5. The consequence was this slight but audible
time-difference resulting often in the pedal being heard to play first.

The roll speed on replay is quite accurate if suitably adjusted with the aid of test rolls. Still, it
was not always perfect. The correct adjustment of a Welte pneumatic motor was possible, but
not simple, and highly prone to small sensitivities, since the technology could become flawed
when worn and lead to erratic results. In any case the operator of the instrument, as Welte’s
own advertising encouraged, could vary the tempo with a lever according to his own whims.
He could also, if he wanted, choose and change the registration. Likewise fluctuations in
speed could be influenced by changing spool diameters as the rolls played.

Furthermore the different specifications of various Welte Models were such that the smaller
organs, and there were many of these, simply could not re-create the detail of the original
registrations for lack of available tonal resources.

On their part the organists - especially when they played works of Reger - had to adapt 3-
manual music to just 2 manuals in Freiburg. This also occurred in a room with rather dry
acoustics, which could not be compared to Cathedrals, for which most of this music was best
suited. So far research has not turned up anything specific, but judging from the rolls, the
organists do not appear to have made significant adjustments for this environment. Gigout’s
recording of his own Grand Choeur Dialogué appears not to have taken the Freiburg studio
acoustics into account.

[01 Gigout 1.wav]

This sounds much as we would expect it to if we simulate a more lively acoustic

electronically.

[02 Gigout 2.wav]

Copies of the original rolls were musically and mechanically "revised" and then reproduced
using a high-speed perforator. However the copied rolls had tiny imprecisions and are
demonstrably not identical. Particularly with regard to the anticipation of pedal entries, and
perhaps also on account of such small vagrancies of reproduction, when the American
 5

organist Thomas Murray asked his teacher Clarence Mader for advice about recording around
1950 he was told: “Don’t bother about the Welte, it breaks chords”6.

Swell crescendi and diminuendi were likewise a compromise. The rolls recorded only closed,
open, slow crescendo/diminuendo (4-6 seconds?) and fast crescendo/dinimuendo (1-2
seconds?). The roll perforations, in theory, could change this instantaneously from any one to
the other. The system was well thought out but it made no absolutely precise and reliable
reproduction of the organist’s swell-manipulations possible. Differing momentums of the
shutters, or the condition of their motors, brought new and differing factors into each
equation.

Of far greater significance was the consideration that the swell enclosed the whole organ. No
complete division was independent of this one enclosure, save the separate dedicated swell
box, as in Seewen, for the Vox Humana. This was not exactly what most of the repertoire was
predicated on in this connection.

Two test rolls are preserved in the Seewen collection. When played they can be used to
assess the correct adjustment and functioning of the organ and its roll-playing mechanisms.
Two examples of these tests follow. In the first the operation of the swell-box is checked:

[03 test roll swell.wav]

So far so good. The second example checks note-repetitions. Some notes clearly fail the test
here. Problems of this kind have serious consequences if phrasing or articulation are to be
based on rolls played under these conditions.

[04 test roll repeated notes.wav]

The roll containing the complete recording of the Suite Gothique of Boëllmann was played
twice for the Radio recordings. The first time there was no tempo adjustment, but for the
second it was sped up. The roll used was recorded by the Swiss organist Paul Hindermann
and is identified as Welte number 752 (undated). Hindermann was born in Zürich in 1868,
studied with Rheinberger, later filled a professorship in Zürich and died there in 1925. Welte
released rolls of his playing in 1912, 1913 and 1926 (the year after his death) including works
by Bach, Brahms, Saint-Saëns, Franck, Boëllmann, Schumann, Guilmant, Salomé and Reger.

[05 Hindermann Boellman normal.wav]

In the next example the complete Suite takes 13'30" to play rather than 15'42", that is around
16% faster. Human perception for such tempo fluctuations is around 5%7. The musical
consequences of this accelerated tempo are immediately obvious. Besides this, however, is
also a change in registration. There could be a variety of explanations for this, such as
technical faults, or the intrusion of an operator.

[06 Hindermann Boellman faster.wav]

Other problems also arise here. Karl Matthaei, 1897-1960, was an important Swiss organist
 6

who, already during the 1920s, pioneered early music8. He recorded rolls for Welte, released
in 1926-7, with a repertoire including Bach, Buxtehude, Scheidt, Praetorius, Sweelinck and
Hanff. The absence of the pedal part in this recording is puzzling and it is hard to imagine
that a person of the stature of Matthaei had not noticed it. The wrong notes were possibly his
own doing, but inexplicably remain unedited. The technology was there to edit them out. It
could also arise from a technical problem of the organ or its roll-player.

[07 Matthaei Christe du Lamm.wav]

Welte technology, even when new, had replay problems, but when it aged and became worn
with use, and particularly from the roll-copying processes, it can only be used with caution as
a basis to determine performance practice. In spite of that we know that all details were put
down on the rolls at the time of recording. It is really only the copied rolls and the replay
technology that cause us problems.

At this point I would like to take a closer look at the work of the Boston organ-builder,
Nelson Barden & Associates. Barden believed that these problems are all redeemable and,
together with others, developed a computer program that resolved the playback limitations of
these rolls. He concerned himself principally with the recordings of Edwin Lemare. In this
connection he decided to reproduce the Welte rolls of Lemare in corrected versions9.

Here are two excerpts from an improvisation by Lemare. Both recordings used the identical
roll, Welte number 1195. Even the playback tempo is almost identical, with less than a 10-
second difference in duration of 7'48" i.e. under 2%. We first hear the Seewen Radio
recording which might be dubbed as "uncorrected Welte Technology." The replay is on an
instrument showing the traces of nearly 3/4 century’s wear and tear. Note particularly the
Pedal and the irregularities in the bell register.

[08 Lemare Improvisation Seewen.wav]

Now Barden’s corrected version10.

[09 Lemare Improvisation - Barden.wav]

Finally an excerpt from Lemare’s performance of his own "Rondo Capriccio: A Study in
Accents" opus 64, corrected and recorded by Barden from Welte Roll number 1181 dated
1913. It demonstrates the technical possibilities that are available to us today, since this piece
represents a significant challenge for player organs.

[10 Lemare study in accents.wav]

The work of Barden & Associates is not merely an invaluable help for future research in the
ambit of performance practice but it also lays the groundwork for commercial productions of
good clean recordings of these famous organists. Apart from that we now also have the
potential to retrieve a world musical heritage that is threatening to decay, but which could be
permanently preserved through these now-available technical means.
 7

oooOOOooo

With the following musical examples we shall have to content ourselves for the moment with
the limitations of the Welte system and an aging organ.

The Seewen collection contains numerous works of Reger. In 1980 Swiss Radio DRS
recorded the following (grouped according to the organists) -

Clarence Eddy
Pastorale in F,Op.59,No.2
Walther Fischer
Gloria Op 59 Nr 8
Mariä Wiegenlied,Op.76,No.52
Intermezzo, Op.80,No.10
Invocation a.d.II.Orgelsonate,Op.60
Kurt Grosse
Toccata und Fuge
Fantasie u. Fuge über B A C H,Op. 46
Fantasia f.Orgel Wachet auf ruft uns die Stimme, Op.52, No.2
P. Hindermann
Ave Maria
Joseph Messner
Romanze, A moll
Consolation Op.65
Franz Philipp
Weihnachten Op.145,No.3
Günther Ramin
Toccata und Fuge D moll, Op.129
Max Reger
Fuge in G (56.3)
Jesus meine Zuversicht Op 67 Nr 20
Wie wohl ist mir, O freund der Seelen Op 67 Nr 50
Lobt Gott, Ihr Christen alle gleich Op.67,No.23
Moment musical,Op.69, No.4
Romance, Op.69, No.8
Melodia, Op.59, No.11
Präludium, Op.85, No.3
Basso Ostinato,Op.92, Nr.4
O Welt, ich muss dich lassen,Op.67,No.33
Benedictus, Op.59,No.9
Wer nur den lieben Gott lässt walten Op.67, No.45
Canzone, Op.65,No.9
Oh, wie selig, Op.67, No. 52
Alfred Sittard
Benedictus,Op.59,No.9Eddy, Fischer, Grosse, Ramin, Sittard and Reger are all closely
connected with the Berlin Organ School of the late 19th and early 20th centuries. A connection
is also established through Franz Philipp since he studied in Basel with a former Straube
 8

student, Adolf Hamm. Philipp was better known as a composer who, incidentally, took Anton
Bruckner as his role-model.

oooOOOooo

In the Seewen roll collection there is not one recording of Reger by Karl Straube. Of
Straube’s recordings only 7 rolls are preserved at Seewen, dated from 1922-1928. These are
recordings of Bach - Chorale-Preludes, mainly from the Little Organ Book, and his Fantasia
in g minor - as well as Buxtehude’s g-minor Praeludium.

Listen now to the most Reger-like of Bach’s works, the g minor Fantasia. Straube’s
performance possesses all the qualities that we would expect, a solid legato and some
crescendi that might appear moderate for a fully-enclosed organ. Note the use of manual 16's
and the constant registration using reeds. It is to be assumed that this recording took place
around the time or soon after Straube had been influenced by the Organ Reform Movement11.

[11 Straube Bach Fantasia.wav]

To simplify matters I have used as my point of departure an assumption that Reger’s

published music presents us with enough detailed markings with which to basically assess the
quality of any performance of it. Against this we know that his metronome markings may not
always be taken as absolute criteria: but they may be read comparatively - also through other
markings such as sempre stringendo etc.. For each of the following pieces I have prepared
basic graphical representations that show the overall dynamic and tempo indications as Reger
indicated them as well as the values the interpreters gave them in these recorded
performances. Naturally it is difficult to give exact numerical parameters to such indications
as "Langsam, doch nicht schleppend". The graphs give only average values for each bar. I
have made no attempt to represent phrasing and articulation in this way. Occasional
observations can be made regarding how each artist details these aspects individually. This
will also be evident from following the scores.

It is equally important to observe when the individual artist omits to follow the directions in
the score, and thus endows the music with his own creative ideas, and to see what they
sometimes do when nothing is marked.

oooOOOooo

Günther Ramin was born in 1898 in Karlsruhe, attended the Leipzig Thomas School from
1910 onwards, then studied organ with Teichmüller and Straube as his teachers. He was
called to be the Thomaskirche organist in 1918 and as such made Welte rolls, dated 1922-6
(Bach, Buxtehude, Händel, Lübeck, Reger and an improvisation on “Vom Himmel hoch”).
He died in 1956 in Leipzig.

From Welte’s roll, no. 1991, we hear his recording of the Toccata and Fugue Opus 129. To
all intents and purposes he remains quite true to Reger’s score. One hears how Ramin allows
some small freedoms to creep in, for example the rhythm right at the outset, which is a very
free reading of the score. However, all in all, it is a performance which follows the detail of
 9

Reger’s intent quite closely. Of interest in the fugue, which is not an accelerando fugue, is
how Ramin nevertheless does introduce some tempo-changes rather along the lines of a
crescendo-accelerando Fugue.

Toccata Op 129 #1

Tempo
Ramin; Noten

Dynamics
Music

Recording

[12 Ramin Reger Op 129 1.wav]

Fuga 129 #2

Tempo
Music; Ramin
 10

Dynamics
Music

Recording

[13 Ramin Reger Op 129 2.wav]

oooOOOooo

Clarence Eddy received part of his education in Berlin. He was born in 1851 in Greenfield,
Massachusets USA, and was a student of Dudley Buck before moving to Berlin where he
studied from 1871 to 1873 with Karl August Haupt and Carl Albert Loeschhorn. He died in
Chicago in 1937.

Considering phrasing, tempo and dynamics, Eddy’s recording of the Reger Pastorale op 59
Nr. 2 on Welte roll no. 1664 is perhaps one of the most true-to-score performances that we
possess of a Reger work. Nonetheless it sometimes deviates clearly from the printed page,
particularly with rhythmic alteration around the main cadences. Waiting for thematic entries
or beginnings of new sections in this way seems to have been fairly common in these roll
recordings. It is, incidentally, assuredly the artist and not a technological idiosyncracy.

Pastorale

Tempo
Music; Eddy

Dynamics
Music
 11

Recording

[14 Eddy Reger Pastorale.wav]

(Certain technical peculiarities in connection with resonant frequencies or aging magnetic

tape etc. have influenced the graphic representation here - in fact Eddy holds his performance
far better to Reger’s dynamic markings than might appear from this.)

oooOOOooo

Reger’s Benedictus is said to be his best-known work. It is thus instructive to compare two
performances of it here. The first is by Reger himself, the second by Alfred Sittard.

It is sometimes claimed that Reger was not a good organist. Born in Bavaria in 1873 he was
already deputizing at Weiden Cathedral 1886-9 - and (according to Groves12) was playing a
repertoire there which included works by Mendelssohn, Schumann und Liszt (it had short-
octave so this may well have been quite an interesting achievement). Later he had a teaching
position from 1905-6 in Munich, which included organ students. So by any yardstick of his
own epoch he was clearly acknowledged as an accomplished organist.

From 1907 on he was Musical Director at Leipzig University and taught composition there.
His Welte rolls are mostly undated, but those that are dated all bear the year 1913.
Presumably if Reger gave up organ playing when he moved to Leipzig then he was possibly
about 5 years out of practice when he made these recordings. The fact that he chose relatively
simple pieces of his own to record may be connected with this. We should not forget that in
the interim he was still making appearances as a pianist.

That Reger recorded anything at all shows how highly his playing was prized. For Welte this
was no altruistic gesture towards posterity but primarily a hard-headed business consideration.
Peter Hagmann13 believes that the well-known recordings of Reger from the 1960s were made
with an apparatus which was running about 20% too slowly. These were also criticized
because manuals and pedals were not sounding together. But as we have already seen, this
could have been a roll-related technical problem. It could equally have been Reger’s manner
of playing14. Hans Klotz noted with some surprise Reger’s use of mixtures, a sign that our
modern concepts of playing traditions in that era can also be erroneous.

With this background let us listen to Reger’s recording of his Benedictus. He remains true to
most of his own playing instructions.
 12

The apparent misreading of two quavers as crotchets in the second section is worthy of
comment. Is it a technical problem, or an errant intervention of one of the company’s editors
during the recording phase, or is this once again merely a different performance criterion that
was unconcerned with wrong notes?

Benedictus

Tempo
Music; Sittard; Reger

Dynamics
Music

Reger

Sittard

Alfred Sittard was born in Stuttgart in 1878, studied in Cologne, became the Dresden
Kreuzkirche organist in 1903, then in 1912 organist at the Michaeliskirche in Hamburg and
in 1925 Professor for organ in Berlin, where he died in 1942. By comparison with Reger’s
recording of this work we find many performance elements that deviate substantially from the
composer’s printed page. Even so, Sittard was a Welte artist who could eventually, of course,
look back on more roll-recordings than Reger. Since Reger died in 1916 he could only play in
what were the early years of this technology. Sittard’s dated rolls are between 1913 and 1924
and demonstrate a serious involvement with the organ repertoire, including Bach, Franck,
 13

Händel, Liszt, Reger, Saint-Saëns and one of his own pieces.

[15 Reger Reger Benedictus.wav]

[16 Sittard Reger Benedictus.wav]

oooOOOooo

Kurt Grosse was a Berliner through and through. Born there in 1890, educated there, worked
there and, as far as we know, died there. He was a student at the “Royal Berlin School of
Music” from 1914-19, worked as organist at the Garrison church in Spandau and after 1920
moved across as organist and choirmaster to the Friedrich-Werder church. Between 1915 and
1928 Welte made at least 57 of his organ rolls available for purchase15. Three important
Reger pieces were amongst the offered repertoire: Toccata and Fuge (d minor/D major),
Fantasie and Fuge on B-A-C-H and the Fantasie on "Wachet auf ruft uns die Stimme".

The BACH Fantasia is played exactly as we might expect, as a Fantasia, closely following all
the rhythmic freedoms that are in Reger’s printed intentions. The Fantasia is not suited to a
graphic representation on account of these complex changes.

[17 Grosse Reger BACH Fantasia.wav]

We then encounter a typical Reger “crescendo/accelerando” fugue. It is instructive to look at

this aspect closely. The crescendo of the recording follows Reger’s printed scheme very
faithfully. What is more remarkable here is how closely Grosse follows the Reger metronome
markings. As with Eddy, there are freedoms around the cadential points and new entries. But
the considerable freedoms that Grosse allows himself on the final page are really quite daring
for all that they are convincing - and also cannot be adequately represented graphically. If we
watch the beats closely here we might recall the momentary quaver and crotchet deviation in
Reger’s playing of his own Benedictus. Again we have to understand this as either a mistake
or interpretation idiosyncracy of the performer. In some bars we might also detect possible
rubato playing, always assuming this is not a technical problem with the rolls.

BACH Fuga

Tempo
Music; Grosse
 14

Dynamics
Music

Grosse

[18 Grosse Reger BACH Fuga.wav]

oooOOOooo

Rolls can unquestionably be used to yield important criteria for performance practice
considerations; but they are most valuable only when the recording and playback technology
is corrected.

However, what we learn from this is not always what we might want to.

From the Seewen holdings of Joseph Bonnet’s roll recordings we observe that, for the Bach
Passacaglia, an organo pleno is used. Pleno stands in one of the early manuscript copies of the
work. The danger here is that we might try to see and use a logical performance paradigm.
This is probably not the case. The less logical such paradigms are, the less we can expect to
find performances on which we can base our own.

It is precisely here that we learn something from these recordings: the performance paradigms
the organists used here appear illogical only to us, not to them. Tuned percussion
registrations, such as bell (Glocken) and harp (Harfe) were not unique to Lemare. Assuming
no Welte-technologiy playback errors, Bonnet also used these registers - quite logically - in
his own Angelus du Soir. Nater used them in Widor’s Symphony VI, in both 2nd and the
famous 5th (sic!) movement and in a Noël of Dubois. As we have heard, Ramin used them in
Reger’s d-minor Toccata. Messner also used them in the Reger Romanze op. 69 Nr. 8 and
Fischer in Liszt’s Weinen Klagen. Bach specified them in Mühlhausen16. Medieval organs
were frequently portrayed together with bells, as in the Rutland Psalter - in fact it is difficult
to believe that organ and bell combinations are not one of the oldest demonstrable ensembles
in music history.

Why do we so rarely hear this combination today?

 15

These recordings also, therefore, encourage us to question our pre-conceived ideas about
registration and performance practice, including questions about the acceptability of wrong
notes. Investigation into the use of the tremulant could also turn up interesting new insights
here.

oooOOOooo

I am delighted to say that the Seewen organ will soon be restored. This is a project involving
a large and extremely rare Welte-Philharmonie organ combined with its own unique and
comprehensive collection of priceless rolls.

Today, almost a century after the introduction of the Welte Autograph organ we stand on the
threshold of preserving these rolls, which are currently in a rapidly-disintegrating paper
medium, through the refined digital electronic systems of our own advanced technologies.
This, quite apart from the all-important question of the preservation of a totally unique and
important musical heritage, also has the advantage, through transcription, technical correction
and publishing of recordings which can now be heard as they always should have been17.

oooOOOooo

First published in German (without sound samples) in Questand II (journal of the Anton
Bruckner Private University, Linz, Austria) © 2006 ConBrio Verlag Regensburg.

For a German version of this paper version please see:

http://galifrey.triode.net.au/~rumsey/linz_reger_deutsch/

© / All rights reserved / Alle Rechte vorbehalten

oooOOOooo

Revision of 23 May, 2011

 16

Acknowledgements

Dr. Christoph Haenggi, Leiter des Museums für Musikautomaten (Seewen, Switzerland)
Dr. Rainer Kaiser (Mauchen, Germany)
Dr. Dean Billmeyer (University of Minnesota, USA)
Dr. Christopher Anderson (University of Buffalo, USA)
Prof. Brett Leighton (Bruckner Universität, Linz, Austria)
Nelson Barden (Boston, USA)
Bernhard Prisi (Seewen, Switzerland)
Elizabeth Rumsey (Sydney/Australia, Basel/Switzerland)

Bibliography and Sources

Weiss, Heinrich Früh biegt sich, was ein Haken werden will (Basel: F. Reinhardt, 1998) pp.
110-120. ISBN 3-7245-1012-8.

Th. Kuhn SA Männedorf Bericht zur Welte-Philharmonie-Orgel, (unpublished).

Verena Gäumann Karl Matthaei 1877-1960: Leben und Werk eines Schweizer Organisten.
Ed. Dominik Sackmann (Florian Noetzel). ISBN 3-7959-0715-2

Kurt Binninger Die Welte-Philharmonie-Orgel in Acta Organologica Band 19 (Merseburger

1987)

Peter Hagmann Das Welte-Mignon-Klavier, die Welte-Philharmonie-Orgel und die Anfänge

der Reproduktion von Musik, Peter Lang Verlag, Bern, 1984.

Q. David Bowers, Encyclopedia of Automatic Musical Instruments, Vestal Press, New York
ISBN 0-911572-08-2 - p. 327.

Nelson Barden Edwin H. Lemare in The American Organist (1986 Vol.20. Nos. 1, 3, 6, 8.)

Database of Seewen Organ Rolls (compiled by David Rumsey 2002)

[Recording] Max Reger spielt eigene Orgelwerke by the Electrola Co. of Cologne, Germany
(1961: 1C 053-28925) on the Welte organ in Wipperfürth when it was acquired by Dr. Weiss
and immediately before it was moved to Seewen.

[CD Recordings] EMI 5CD set 7243 5 74866 2 0 CD 2 (Reger amongst others, recorded off
the Welte organ in Linz am Rhein, Germany.

[In print] David Rumsey Organists on a roll - the Welte organ's mechanically-recorded
performances. Conference Paper 2002 (Arizona). Due for publication 2005 by GOART
(Göteborg, Sweden).
 17

Appendix

In the Welte premises at Freiburg (Breisgau, Germany) there were two organs side-by-side,
one for recording and the other for playing back. They no longer exist but apparently had
specifications as given below18.
The exact specification of the recording organ is not fully clear. It appears that an
instrument built in 1909 was enlarged somewhat in about 1913. The details of these
instruments, so far as we can be sure, are sufficiently interesting to warrant their both being
given here.
1909
(After Kurt Binninger in Acta Organologica 1987 Vol Bd.19)
Manual I Cello10 m 8
Principal1 m 8 Posaune w or m 16
Traversflöte2 w 8 Manual II
Viol d'orchestre3 m 8 Viola11 m 8
4 12
Gamba m 8 Wienerflöte w 8
Vox coelestis5 m 8 Aeoline13 m 8
6 14
Fagott papier-mâché 8 Bourdon stopped w 8
Flöte7 w 4 Horn15 w g0 - 8
Harfe8 m (plates) G- Klarinette16 papier-mâché 16
9
Glocken m (tubes) Oboe17 g0 - 8
18
Pedal Posaune C-f -<0
8
Violon open w 16 Trompete19 g0 - 8
20
Subbaß stopped w 16 Vox humana m 8

Compass: Man I: C-a3; Man II: C-g3 (sic!); Ped: C-f1; Action: Pneumatic (pouch/membrane);
Location: To the right of the playback organ. Both organs totally enclosed in a single swell-
box; Pipework:
1
"gentle singing tone"
2
"internal" lips C-f<0; g0- overblowing, typical flute tone;
3
"beautiful string-tone"
4
"stronger than the viol d'orchestre"
5
"Keen tone; tuned as a beating rank to be used with the viol d'orchestre"
6
C-b0 free reed with long wooden boots and covered resonators, very similar to the
orchestral insrtument; c1- flue pipes with Fugara scaling.
7
C-b0 resonators tapering front to back; c1- very wide scaling giving a round flute tone.
8
metal plates placed over wood or papier-mâché resonators and hit with pneumatic hammers.
9
C-g0 metal tubes, sounded as per Harfe.
10
often borrowed from the Violon 16.
11
tonally very similar to a Geigenprinzipal
12
harmonic flute, bass only (not harmonic in the trebles) scaled 2 semitones narrower than
Traversflöte in Manual I.
13
somewhat more gently voiced than the Man I Viol d'orchestre
14
voiced to sound full and round
15
Flue rank of special construction - as pictured in Acta Organologica - but was not
harmonic, possessing a very carrying tone;
16
free reed, similar to the Fagott, but with wide-scaled, open resonators. Very similar to the
 18

characteristics of the orchestral instrument.

17
Plays from C in the Tutti. Beating reed full-length resonators, lengthened tops and turning
caps.
18
conical tops
19
continuation of the Posaune
20
built as a Silbermann Vox Humana
c1913
Manual Wienerflöte 8
Bordun 16 Aeoline 8
Principal 8 Bordun 8
Traversflöte 8 Dolce 4
2
Gambe 8 Quinte 2 /3
Viol. d'orch. 8 Clarinette 16
Vox coelestis 8 Trompete 8
Flöte 4 Horn 8
Piccolo 2 Oboe 8
Sesquialter Pedal
Fagott 8 Violonbass 16
Harfe Subbass 16
Glocken Cello 8
Manual Gedackt 8
Viola 8 Posaune 16

Compass: Manuals: C-a3 (c4?); Pedals: C-f1(?); Accessories: Vox Humana Echo (opens/shuts
Vox Humana's separate box lid); Tremolo. Fully enclosed; Action: Pneumatic; Location: In a
spacious but carpeted room in company premises; Console: Mounted on a podium, projecting
forward from the facade, the organist facing the instrument; Pipework: This instrument
possibly inherited much of the 1908 organ’s pipework; Destroyed during an air raid in 1944.

The current specification of the Seewen organ given below - built between 1912 and 1920,
enlarged by Welte in 1937 and 1978 - can here be compared with the above Freiburg
recording/playback organs as well as some Welte organs connected with the USA. The organ
in the New York recording studio possessed 3 manuals and was tonally somewhat differently
endowed. This instrument, and the rolls made in New York, represented a variant musical
world to that associated with Freiburg19. (The New York recording apparatus is also now
preserved in Seewen.) Lloyd Davies was a former employee of the US branch of Welte. He
relayed a "typical Welte specification". Of considerable importance also was the fact that he
noted the relative dynamic power that the ranks had to have - given here in square brackets -
so ordered that the rolls would play with the correct musical dynamic balance. "Germany"
was Davies' recollection of a typical imported Welte organ coming from Freiburg. It
corresponds to one of the smaller of Welte's models20.
 19

Seewen Freiburg NY "Lloyd" "Germany"

Manual I I Manual Manual Manual Manual
Bordun 16 Bordun 16
Principal 8 Principal 8 English Diap Diapason [3] Principal
Traversflöte 8 Traversflöte 8 Concert Flute Fl traverso [2] Fl. Traverso
Gedeckt 8
Gambe 8 Gambe 8 Viola Violin solo [4] Gamba
Viol. d'orch. 8 Viol. d'orch. 8 Dulciana VdO [1] VdO
Vox coelestis 8 Vox coelestis 8 Unda Maris Voix Celeste [2] Voix Celeste
Octave 4
Rohrflöte 4 Flöte 4 Conc. Flute 4 Har. Fl. 4[3] Flute 4
Nachthorn 2 Piccolo 2
Mixtur Sesquialter
Trompete 8
Fagott 8 Fagott 8 French Horn Bassoon [5] Faggot
Harfe Harfe Harp Harp Harp
Glocken1 Glocken
Manual II II Manual Solo Solo Solo
Viola 8 Viola 8 Open Diap Diap 8-4-2 Diapason
Harmoniefl.2 8 Wienerflöte 8 Bourdon Bourdon [2] Bourdon
Aeoline 8 Aeoline 8 Philomela Philomela [3] Weiner Flute
Bordun 8 Bordun 8 Flauto Dolce Flauto Dolce [1]
Blockflöte 4 Dolce 4 Gamba V.d’Gamba [2-3] Viola
Quinte 22/3 Quinte 22 /3 Aeoline Aeoline [1] Aeoline
3
Terz 1 /5 Clarinet 16 Clarinet 16[3] Clarinet 16
Quintzimbel 1 Cornopean b+t Cor Angl b+t [4-5]
Sesquialter Muted/Trpt Trumpet b+t
Clarinette 16 Clarinette 16 English Horn Horn (?labial)
Trompete 8 Trompete 8 Oboe Horn Oboe [3] Oboe
Horn 8 Horn 8 Vox Vox [2-3] Vox
Oboe 8 Oboe 8
Vox humana 8
Glockenspiel 2
Pedal Pedal Pedal Pedal Pedal
Violonbass 16 Violonbass 16 Tibia Minor Bourdon Subbass
Subbass 16 Subbass 16 Viola Violone Violone
Gedackt 16 Cello 8 Concert Fl 8 Cello 8 Cello 8
Gedackt 3
8 Gedackt 8 Tuba Profunda Trombone Posaune
Posaune 16 Posaune 16
[Trompete 8
[Clairon 4
[Sing.Cornett 2
1
1937? Echo
2
original = Wienerflöte Cor de Nuit
3
spielbar nur von
Viola Aetheria
Rollen
Vox
 20

1. Binninger, Kurt, Die Welte-Philharmonie-Orgel, Acta Organologica Vol. 19, Berlin -

Kassel 1987 ISBN 3-87537-227-1

2.The symbol of a smoking chimney in those days did not have the negative connotations of
"environmental pollution", rather that this was a very well-to-do business with the latest
equipment, particularly steam-powered machinery. The house organists were Franz Philipp,
Johannes Diebold, August Heim, C. Hofner and Bernard ten Cate. Kurt Binninger also
recorded rolls for them.

3. For further details see David Rumsey Organists on a roll - the Welte organ's
mechanically-recorded performances. Conference Paper 2002 (Arizona). GOART
(Göteborg, Sweden)

4. Heinrich Weiss, Festschrift on the occasion of the dedication of the organ.

5. The system is actually far more complex than it sounds from this summary explanation.
Depending on which couplers were drawn and which of the 30 lowest notes in the Hauptwerk
and/or Pedal were played, the advances or delays could vary. For more details of how it was
achieved see Kurt Binninger’s article in Acta Otrganologica Op. Cit.. To be noted, however,
is that whenever the second manual to pedal coupler is drawn (or notes on the second manual
play through borrowing or extension) the precise point at which the organist played the
original pedal note can be determined. This is the principle Barden adopted to correct the
manipulated multiplexing of the pedal in the original roll editing and copying processes.

6. Communication from Nelson Barden 2004/5

7. From discussion with Nelson Barden regarding his observations during recording sessions.

8. Verena Gäumann Karl Matthaei 1877-1960: Leben und Werk eines Schweizer Organisten.
Ed. Dominik Sackmann (Florian Noetzel). ISBN 3-7959-0715-2

9.It should be again noted that there was no adjustment forward or backwards of notes from
the coupled second manual. This allows an easy determination in re-establish the exact point
at which the organist originally the pedal note.

10.Some people think the first recording “sounds better.” And that may well be the case from
the recording, acoustic, and other qualities of the Seewen organ as opposed to that in Boston,
or the recording of them. Our perceptions and feelings, as Professor Leighton quoted - from
Hill - are not the prime question here. In spite of acoustic, organ, recording techniques etc.
the original performance of Lemare is far more accurately reproduced in all technical aspects
with the Boston CD - not to speak of the fact that all the notes are actually played and none
are missing.

11. It is to be noted that this performance also deviates from Straube’s own Bach edition.

12. Ed. Stanley Sadie, New Grove Dictionary of Music and Musicians, MacMillan Publishers
Limited, London, ISBN 0-333-23111-2
 21

13. Peter Hagmann Das Welte-Mignon-Klavier, die Welte-Philharmonie-Orgel und die

Anfänge der Reproduktion von Musik, Peter Lang Verlag, Bern, 1984.

14. In that Reger had apparently left the Hauptwerk uncoupled to the Pedals in the first and
last sections he precipitated the maximum sidereal shift in these parts of the piece. Which
means that if we want to emulate this as a performance practice then we need to realize that
this anticipatory manner of playing the pedal was never so great as it might appear because a
component of it was assuredly the technology and not Reger. Apart from this we might also
observe that the effect is most noticeable on the first note of a pedal entry - the later ones are
much closer to a good ensemble. Was he out of practice? Was it an unimportant
performance paradigm to play precisely together? Was he simply drunk? Or could this also
be connected with the fact that these recordings were made relatively early in the history of
Welte (1913) and that the technology - especially that of roll editors - was still in its
development stages?

15. It should be noted here that, for a German of his era - active in Berlin - he also played and
recorded contemporary French works of his own era, e.g. Widor. In this way he singles
himself out from most of his contemporaries.

16. In Smets Die Orgelregister ihr Klang und Gebrauch Rheingold-Verlag Mainz, 1948, 7th
Edition, 1968, under the entry “Glockenspiel” there is a list of many organs with one or more
such registers, including (16.-18. Century): "Hamburg, Jakobikirche; Altenburg,
Schlosskirche; Breslau, St. Maria Magdalena; Erfurter Augustinerkirche; Halberstadt,
Katharinenkirche; Königsberg, Haberbergerkirche and Neustädterkirche; Magdeburg, St.
Ulrich". Then, “after a long time the Glockenspiel reappears towards the end of the 19th
century ... ”: Amongst the "best" of these are: "Berliner Dom and Wilhelm Gedächtniskirche
(Sauer); Braunschweiger Dom (Furtwängler & Hammer 1901); Breslau Festhalle (one each in
2nd and 5th manual); Steinmeyer organs in Augsburg, Stadhalle and Ludwigshafen, Meeting
Hall of I. G. Farben & Co. (Pedal and Fernwerk); Dortmund, Rainoldikirche; Erfurt,
Predigerkirche; Eßlingen, Stadtkirche; Hamburg, Michaeliskirche (one high and one low-
pitched); Ilmenau, Stadkirche; Music Exhibition Frankfurt (Main) 1926, all built by Walcker;
Heidelberg, City Hall (Voit). Furthermore there are reports of bells, including sleigh bells in
America and Spain (the Campanologo respaldo, in Sevilla Dom). One could also add
Sydney Town Hall here.

17. Or, with another mouse-click, the way they sounded then. It is to be stressed here that the
exact playing manner of the organists, whether good or bad, whether together or not - all the
detail of their performances can now be reproduced in spite of the earlier intervention of
Welte’s system or their editors.

18. Kurt Binninger, Op. Cit.

19. The repertoire originating from the New York studio recordings was slanted more
towards theatre and cinema organ music. While Freiburg included these in principle it formed
only a backdrop to the classical repertoire from pre-Bach to post-Franck, Reger and the
"moderns" of that era.
 22

20. The first column is from Seewen, all others are personal communications from Nelson
Barden in Boston at the end of 2004 or beginning of 2005.