Last weekend, I had the chance to visit the owner of a Triola for
whom I have arranged, over the last 18 months, about 15 pieces.
The Triola owner was not really satisfied with my arrangements up to
now; my visit was triggered by two arrangements which he sent back as
"not acceptable".

As a side note, the Triola, numbered 8497, was bought by the father of
the current owner in or around 1923, mainly as a means of earning money
in the then very bad times in Germany.  Thus, the current owner is not
a collector of any sorts -- which appears to be an exceedingly rarity
in our hobby (with the exception of player pianos).  He told me that he
routinely uses the instrument to play at meetings with his colleague
ex-miners in his home area of the Erzgebirge in Saxonia, i.e., this
Triola is still used in "real life" (as opposed to "collector's
demonstrations" or the like).

I have learned much from that visit, which I'd like to share here.

1. The owner must know how to tune the Triola.

I got a recording from the Triola where I heard a C-sharp in a few
arrangements in C or G major, an impossible note.  I thought that the
note was stuck; however, it turned out that it was simply a not-tuned
D-string which I needed as 5th of the G major chord.

The owner had already purchased a Korg tuner, and in a few minutes we
had the Triola roughly in tune, with the owner being comfortable with
the use of the tuner.  Incidentally, we left the tuning of most strings
as they were, which resulted in a=425 Hz.

2. Regulation of the plucking springs

An additional problem with the aforementioned D string was that it
sounded much more prominently than all the other strings.  I did a
complete re-regulation of all 25 springs (turning the small screws),
as follows:

* Putting the forte-piano lever to piano.

* For each note, taping all the slots in the "key guide block" (or what
it might be called) except the one to be regulated; and then regulating
the spring until it sounded very faintly.  (The owner, watching me,
stated that this had probably never been done in all the 80 years since
the Triola was purchased.)

A question to Triola specialists: After the regulation, chords sound
very "equalized".  However, we both had the feeling that this sound,
even with mezzo-forte, is already quite harsh; with forte, it's almost
unbearably "metallic".  Is this so?

3. Speed.

I timed the rpm's of the crank when turned by the owner quite a few
times: It was consistently very near to 150 rpm.  This statement must
be qualified as follows:

* The accompaniment of the pieces was new to the owner (I had to change
it on the spot; see 6. below).  He stated (and it was obvious) that he
had to concentrate on the upcoming numbers so much that it was only
possible for him to turn at his "natural speed".

* The owner also stated that with pieces that he had practiced well, he
could change the speed somewhat, but not so much that it would make a
marked difference.

Thus, the "natural speed" most probably varies somewhat with the tunes,
the practice, and the person cranking.

The 150 rpm's result in an initial paper speed of
150 rpm * 25 mm * 3.14 * 8/60 / 60 sec/min = approx. 26 mm/sec
where 25 mm is the diameter of the take-up spool and 8/60 is the ratio
of the gears from the crank shaft to the spool shaft.

(Because of the small diameter of the take-up spool, the build-up during
play has to be considered.  For the paper we use, we arrive at approx.
45 mm diameter after 7200 mm of paper on the take-up spool, or around
48 mm/sec final speed.  The average speed in this case is around
39 mm/sec.  But this "spiral compensation" taken care of by the
punching program I wrote.)

In tendency, I had the arrangements punched somewhat too slow -- the
wife of the owner likened my arrangement of the (fast) Petersburger
Schlittenfahrt [Petersburg Sleigh Ride - galop] by Johann Strauss,
Jr., with a funeral march...  I will, in the future, rather go for
a somewhat "aggressive tempo", probably assuming an initial speed of
23 or 24 mm/sec.

4. Minimal note length

The minimal length of a slot seems to be around 7 mm.  Even with this
length, we sometimes had notes not sounding, but the reason might be
in the plastic-enforced paper we use: sometimes the last punch of a
slot left the paper in the slot.  I have to check for this more
carefully.

5. Maximum slot length

Because it is a "keyed" instrument, the Triola requires contiguous
slots.  However, because of the use of paper, the maximum slot length
should not be too long: around 100 mm, maybe 120 mm.  Thus, it is
necessary to

* either add additional bridges when a note becomes longer;

* or limit the length of each note in the arrangement so that the slot
lengths remain below the threshold.  This is especially painful -
because of the spiral compensation - towards the end of a roll.

To my astonishment, I found that bridges of 1 mm length are virtually
indiscernible in the "vibrato" sound of the Triola.  This simplifies
arranging considerably.  However, see also 7. below.

(For the plastic-enforced and virtually untearable paper we use,
a 1 mm bridge is sufficient.  This might be quite different for other
paper sorts.)

6. Possible harmonics of the bass strings and string groups.

This is probably the most important finding for my arranging.  Let me
describe my mistake first:

The stamped bass numbers allow either chords or single bass strings to
be played.  I used the latter possibility extensively for more lively
bass lines.  Already for a simple march, I would notate for the initial
phrase, which is totally in G: (2) (4) (2) (4) etc., assuming that
the notes that would sound are G D G D.  The harmonic meaning should
have been <main note of G-chord>; <fifth of G(!)-chord>, etc.

However, in practice, the Triola player, when instructed to pluck
_only_ the D string, will afterwards slightly -- or not so slightly --
touch the next string to the left, which is F-sharp!  Thus, in that
purely G phrase, we always hear a very very disturbing F-sharp.

In effect, this means (and I think Nicholas Simons has already pointed
that out more than 18 months ago, but I did not understand it then)
that also single bass notes are essentially only usable for the straight
major chords accompanying them.  In other words, the variability of the
bass lines is much more constrained than a "theoretical view" might
imply.

During my visit, when I understood this problem, I wrote new numbers
on the initial part of some rolls and had them played.  The result was
a virtual disappearance of "strange wrong notes".

However, I think there is still some leeway if one looks carefully at
the bass chords.  Here are the bass strings and the first accompaniment
string to the left of each bass string, together with a list of
possible harmonics:

1  C G   C major, c minor
2  G G   G major, G minor; but also fifth of C major/minor; and
even third of E minor, E-flat major; and septieme of A major.  This is
the only bass string that can be used as I did it with all of them.
3  F A   F major
4  D F#  D major
5  A C#  A major
6  E G#  E major

For bass number 3 to 6, it is really impossible to "re-use" the single
bass strings for different harmonics.  But numbers 1 and especially 2
allow for some variation.

I have not yet tried to arrange complete bass lines according to this
table - but I'll try it in the next weeks so that "my" Triola owner
finally gets acceptable (or even good?) arrangements.

7. Countermelody and ornaments.

Last, but not least, the use of countermelody and ornaments must be
quite restricted in a Triola arrangement.

Technically, the reason is that in contrast to crank organs (my
predominant area of arranging), the plucked strings seem to have so
many harmonic tones that too many notes completely hide the melody.
Also, the very important aural information of tone beginnings vanishes
below the "constant fire of re-plucking".  Thus,

* I have to avoid ornaments above the melody line completely;

* I must reduce the customary long notes of a counter-melody to short
plucks (which also alleviates the maximum-slot problem from item 5.
above).

* I must withdraw some of the more extreme jazz chords I like to use,
containing seconds or fourths or sevenths (and sometimes all of them...).
This is especially a pity as the Triola has two complete chromatic
octaves in the melody strings, which in theory allow for all possible
chords -- but only in theory.

As a final consequence, this means that I will have to refuse to
arrange some melodies which I initially thought could be arranged for
Triola.

That is what I learned about the Triola a few days ago.  Maybe it is of
interest also for others.  For me, it means that I will re-work all the
arrangements I did; and of course replace all the unplayable rolls with
newly arranged and newly punched and newly stamped versions for free...

(Side remark: The most boring part is the stamping of the numbers;
fortunately(?), in my job as a software consultant, I have to live
through enough "hotel evenings" with ample time to do such jobs.)

Finally, I must thank very very very much "my" Triola owner from
Saxonia for bearing with the problems described here for quite a long
time, and for giving me the chance to understand how a Triola is really
used by an ambitious layman player for the purpose for which it was
invented -- to entertain friends with a piece of music.

Regards

Harald M. Mueller
Grafing b. Muenchen
Germany

 [ Many interesting Triola articles are indexed at
 [ http://mmd.foxtail.com/Archives/KWIC/T/triola.html
 [ -- Robbie