Dear MMD:

I can certainly agree with the daunting warnings about the difficulty
of building your own musical box. I have been working on exactly that
project for the last two and a half years. In summary I can agree with
the conclusion of other correspondents: no it is not impossible; yes
it is a very big job. I can also agree with Larry Smith; it is more of
a career move than a hobby. I am starting a small business making
accurate reproductions of historic instruments. I could not imagine
having the time and equipment resources start such a project with any
less of a commitment.

My starting point is the production of a 13 inch (12 pouce) Longue
Marche box with an eight tune operatic program. I will be bringing one
to the MBSI annual meeting in Chicago so look for me there: Herz Music
Box Co.

I can offer the following comments on the process:

Most of the mechanism is fussy but not overly daunting for a good
machinist. The governor has more to do with clock making than
conventional machine shop practice, so a nodding acquaintance with that
specialty would be useful. The ability to cut your own gears is very
useful, but not strictly a necessity - you can get commercial gears
that are close to a good match. Of course the advantage to building
from scratch is that there is no need to try to fudge a gear of non-
standard pitch in order to match an existing gear.

It is daunting to know that after you have produced the bedplate,
arbor, governor, spring motor and change mechanism all you have is what
the old makers dismissively referred to as a "blank". The heart and
soul of the movement is the   cylinder and comb. Both of which are
VERY difficult, especially on this scale (92 teeth, 8 tunes,
736 tracks).  Since this is a commercial endeavor, it is not possible
for me to prick and drill by hand. However, over the course of this
work I have thought a lot about how cylinder pricking was done at the
cottage industry level. I think that the prick marks in the dead band
between the beginning and end of tunes on the cylinder offer a clue
The cylinder prickers were good, but they were not magicians. I fee
certain that they pricked the cylinders on some sort of bench arbor
with a guide to rest the center punch on and a clock face on one end to
measure the rotation. My guess is that they had a horizontal punch
rest running axially against the cylinder, and a number of adjustable
stops which they could place along this guide. It appears to me that
the cylinders were scribed with the combs themselves, assuring that
the scribe lines reflected the exact location of the tooth tips. The
cylinder is rotated until the guide is at the center of the dead band.
The pricker punches a mark at the center of one of the tooth grooves.
These punch marks are still in evidence on old cylinders. With the
punch resting in this prick mark, the adjustable stop is brought up
against the side of the punch.  Thereafter, any pin on that track can
be punched by setting the rotation of the cylinder as indicated on the
paper score. The punch is rested on the horizontal rest and pushed
against the adjustable stop to make the punch mark. The result is the
ability to punch any pin hole for a given tune. To prick other tunes
require that the guide moved to the left or that the cylinder be moved
to the right. My guess is that the snail cam that was to be used on
the individual box was used to move the cylinder on its arbor, thus
assuring that the track spacing of the cylinder exactly matches the
step height of the lands on the worm.

This explanation is getting a bit long-winded, but I believe it to be
the method that was used. The bench arbor could be fairly easily
produced in small volumes for the various home workers who did the
punching yet the process could be reasonably accurate.

Not having a retinue of cottage workers myself, I have been obliged to
use more elaborate technology. I have developed a computerized
cylinder drilling and pinning machine. I have used it to read the pin
location of an existing cylinder and reproduce it on a new cylinder. I
can also pin a new work fairly readily, especially if the data is in
digital form: either a list of pin locations or a midi file. Of course
I would also need the comb tuning. As has been pointed out elsewhere,
choosing a comb tuning is not easy and has to be done at the same time
as the arranging of the composition. You could arrange a composition
without taking care to rely on just a few pitches for sustained notes
or trills, but the result would either require a comb with a vast
number of teeth or a comb with a very narrow range, and the ability to
play five or six octaves is one of the great pleasures of arranging for
a large box.

I would be happy to talk with anyone interested in purchasing a
movement which plays a tune or tunes that they have provided. It would
be a very interesting project. It would not necessarily be a great
cost savings over requesting a custom Rouge movement. I expect my
movements, in a traditional case with marquetry lid to be available
for something under $5000. If the number of teeth were the same as my
current combs and the note range of the comb similar plus or minus
half an octave, then I would not need to machine a single custom comb.
The cost of programming and producing a new cylinder would not be too
enormous.  Nevertheless, the total cost may well be comparable to
ordering a custom run of small Rouge movements. On the other hand,
one 92 tooth musical box may be of greater interested to a composer
than 50 cheap 18 note movements.

I will be writing more when my musical boxes are available for sale.
In the mean time anyone interested in discussing the production of a
custom box can contact me at:
  jeherz@sover.net.geentroep [drop .geentroep to use ] or

Jonathan Herz
Herz Music Box Co.
2 Loomis St
Montpelier VT
05602
Phone: (802) 223-5049
Fax: (802) 223-8982

See you in Chicago

Regards, Jonathan Herz